Automation and the Future of Work (2020) by Aaron Benanav

​Quite a landmark study in the literature that redirects the current narratives on automation into a new direction. I’m not sure how the “automation theorists” (as Benanav calls them) will respond to his study. I’m sure the academics will be forced to engage, while the consultants and pop theorists will continue to spew brand-market content (“research”).
 
I begin with a brief summary of the automation discourse, although it has, in recent years, been widely propagated in the public sphere: technological developments are disrupting the labour market, displacing workers, and threaten to (or promise to, depending on ideological inclinations and positionality) automate a large portion of jobs. For some writers in the field, these technological developments are exponential (e.g., Kurzweil), promising vertiginous changes in the near future. The outcomes will disrupt the basic functioning of capitalism, leading to “a new form of life that does not organize itself around wage work and monetary exchange” (p. 7).
 
Supported by data and rigorous scholarship, Benanav shows that the cause of job loss is the lack of growth in productivity, not the uncontrollable growth of productivity that is the result of technological development and automation. Benanav proposes a simple equation to capture the relationship between rates of employment growth, productivity growth, and output growth, which is true by definition: Δ output - Δ productivity = Δ employment, or the rate of growth of output minus the rate of growth of productivity equals the rate of growth of employment.
 
In the narratives of automation theorists, the rise of labour productivity was responsible for negative rates of employment in the manufacturing sector, leading to de-industrialization; in fact, de-industrialization happened due to slowing rates of output beginning in 1973. De-industrialization began as a phenomenon in the Global North and soon spread throughout the world not because of technological growth, but because of worsening overcapacity. In the post-WWII era, the U.S. shared its technologies with countries like Germany and Japan, which pursue export-led national development, leading to overcapacity in the global market, and depressed prices worldwide. In this context, firms responded by globalizing production as more countries competed to enter into the global supply chain. In an interesting reversal, the countries that have suffered less from de-industrialization are those that have robotized more rather than less—high degrees of automation have led to competitive advantage in world markets, helping workers retain their jobs.
 
The manufacturing sector is important due to the sector’s importance in the overall economy. Benanav shows that manufacturing output rates are closely related to the overall national GDP growth rates, given that “in terms of gross output—which unlike value added includes the costs of intermediate inputs (that is, the goods and services consumed by firms)—manufacturing’s “footprint” on the wider economy is significantly larger” (p. 35). The current sluggish growth in the global economy is due to the lack of another sector to replace the manufacturing sector (despite the hype around the ICT sector, I suppose). Overcapacity and decline of manufacturing output has led to a decline in levels of investment, and subsequently, a decline in demand for consumption and reduced levels of hiring. The historically low interest rates in the previous decade and a half have led mostly to financialization, stock buybacks, dividends etc., that characterize financialized capitalism. Due to the growing bubbles of financialized assets, Japanification is a looming threat.
 
With this said, Benanav does acknowledge the potential of technology to affect the demand for labour, but low levels of investment and low costs of labour in a slow economic growth environment make technological automation a secondary cause, and not the primary one. Additionally, in the current ecosystem of technological development, firms incentivized by profit are unlikely to invest in full-automation technologies. Instead, leading tech companies are driven more by the logic of surveillance capitalism—“rather than focus on generating advances in artificial general intelligence, engineers at Facebook spend their time studying slot machines to figure out how to get people addicted to their website, so that they keep coming back to check for notifications, post content, and view advertisements” (p. 40).
 
Theories of automation predict high unemployment rates. But instead of high unemployment rates, there is underemployment and precarity as a larger number of workers crowd into the service sector. The service sector exhibits different characteristics than the manufacturing sector, as there are lower rates of productivity growth and less opportunities for expansion. In fact, only by industrializing the sector has productivity increased, demonstrating the importance of the manufacturing sector once more. The service sector is a stagnant sector in the economy, reliant on the growth the overall economy. And without productivity growth, service sector workers are forced to accept suppressed wages to work, unless there is concerted political action to change the playing field.
 
In light of his diagnosis of the under-demand for labour, which again, is due to industrial overcapacity, de-industrialization, and underinvestment, Benanav re-evaluates some of the popular policy proposals discussed in public discourse. Speaking on Keynesian policy proposals, he notes that, contrary to historical (common sense), “the era of counter-cyclical spending began in earnest in the 1970s” (p. 67); prior to this, the manufacturing sector in the post-war era was already characterized by strong labour demand. (An aside: it may be for this reason that Benanav holds an unorthodox view on financialization, particularly for someone on the left—it seems to me that in his view, financialization occurred due to the lack of productive investment opportunities, and not due to the neo-liberal turn.) He does return to Keynes and his discussion on economic maturity or the “secular stagnation,” in which “it would make more sense to intervene to shrink the labor supply rather than to stimulate labor demand, increasing leisure rather than output” (p. 69). However, this would require socialization of investment levels, as opposed to the current system where private investment can threaten a capital strike. The only way out of this conundrum would be through strong social movements, which is lacking in the current environment.
 
Benanav is also skeptical of UBI as a policy proposal. In a low-growth environment, he fears that UBI will likely take on the right-wing variant that dismantles the remains of the welfare state without reorganizing production. With current environmental problems, one is unable to outgrow the problem of labour surplus—“only a conquest of production, which finally succeeds in wresting the power to control investment decision away from capitalists, hence rendering the capital strike inoperative, can clear the way for us to advance toward a post-scarcity future” (p. 79).
 
In the final chapter, Benanav sketches out his vision of a post-scarcity world of co-operative production, where there are greater individual freedoms in the form of more free time for all. As he notes, “ for a post-scarcity society to come into being, a literal cornucopia is not required… abundance is a social relationship, based on the principle that the means of one’s existence will never be at stake in any of one’s relationships” (p. 89).
 
Additional notes:
 

1. How can economically undeveloped countries pursue development without the manufacturing sector?
2. In the post-COVID-19 world, marked by a decline in globalization and neo-liberal economic orthodoxy, there are debates about repatriating of supply chains and returning manufacturing jobs to the Global North. How might this interact with the problem of over-capacity in global markets on the national level?

 

 

---