What do you think?
Rate this book
Smart Machines and Service Work: Automation in an Age of Stagnation
In recent decades digital devices have reshaped daily life, while tech companies’ stock prices have thrust them to the forefront of the business world. In this rapid, global development, the promise of a new machine age has been accompanied by worries about accelerated joblessness thanks to new forms of automation. Jason E. Smith looks behind the techno-hype to lay out the realities of a period of economic slowdown and expanding debt: low growth rates and an increase of labor-intensive jobs at the bottom of the service sector. He shows how increasing inequality and poor working conditions have led to new forms of workers’ struggles. Ours is less an age of automation, Smith contends, than one in which stagnation is intertwined with class conflict.
192 pages, Hardcover
Published December 23, 2020
301 people want to read
Ratings & Reviews
Friends & Following
Create a free account to discover what your friends think of this book!
Community Reviews
Displaying 1 - 5 of 5 reviews
September 22, 2025
I read this book concurrently with Benanav’s Automation and the Future of Work. At first sight, they are pretty similar, covering the topic of automation with a Marxist/leftist perspective. But there are important differences between both, and mostly in Smith´s favor, at least from my perspective.
Benanav, as I discussed in my review of his book, has many issues in the way he treats data, using imperfect information to sustain his thesis. Smith, on the other hand, does a masterful work of deconstructing the existing data, showing the multiple issues in which our datasets compare between industrial and service work. This critique helps understand how the data is fuzzy, and how the category of service work in itself is a combination of different categories, some productive, some with little influence in production, and some completely separated from production. By using the Marxist notion of productivity, he is able to show that automation is not possible for every category of what we call services.
But this is not all: Smith takes to heart the comments from Marx in the Machinery and Modern Industry chapter of Capital, to show that automation is not, paradoxically, automatic. When automation happens in one sector of the economy, there is less incentive to automate another sector, since the “supply” of labor power increases and thus wages fall, in some cases even lower than the cost of the machinery that would replace them.
This already happened in Marx’s time. In the aforementioned chapter he shows how in England the number of workers was lower than the number of personal servants. Smith takes this idea to talk about a “servant economy”, an economy that is unproductive in Marx’s term, in which most workers have tasks like teaching, elder care, health services, etc that cannot be automated because they involve a human physical/emotional factor. As Marx predicted, with the advance of machines in production, more and more people would fall into these “unproductive” activities, with lower salaries and worse working conditions.
I have broad agreement with this argument, I think it is elegantly argued, and follows closely Marx’s theory. However, the question is once again What is to be Done? Here Smith has an interesting argument. He says that industrial workers have now lost some power, since the threat of offshoring has made Unions more amenable to agreements with Capital. Instead, there are increasing possibilities in, for example, teachers.
This is because teachers are not only (maybe not even primarily) providers of knowledge, but they in fact provide child care, something that was made very clear during the pandemic. I have to say that this argument is interesting but I do not feel it is entirely accurate. First, there is the problem that offshoring has limits, as we can see in the Chinese case. Although for years there have been threats that China would suffer from offshoring, it seems like the supply chain power of the country has proved resilient, at least for the moment. That means that any negative effect on the Chinese industrial supply chain is felt across the world, and there are no easy replacements. This gives Chinese workers in particular a lot of power.
Second, there are some limits to teachers’ action that are related to societal support. Smith notes how in the US, there teachers’ strikes were supported by parents. And that is a key element. If we see, for example, Argentina, the fighting power of teachers has eroded since successive governments have worked hard to break the potential alliance of teachers and parents. Because the main power of teachers is the refusal of provision of child-care, this has a negative effect on parents, who have to juggle care and work. This means that the alliance between both groups rests on tenuous grounds.
That said, these are just some brief objections that I think wouldn’t be rejected by Smith. The point is more that the political and economical power from workers groups is deeply contextual, depending not only on their “place in the production system” but also in circumstantial conditions. In this, Smith is smart to signal that some groups that have been historically forgotten by some Marxists, owing to a misreading of Marx among other reasons, should be reevaluated as workers that could provide the key for the advance of the socialist cause.
Benanav, as I discussed in my review of his book, has many issues in the way he treats data, using imperfect information to sustain his thesis. Smith, on the other hand, does a masterful work of deconstructing the existing data, showing the multiple issues in which our datasets compare between industrial and service work. This critique helps understand how the data is fuzzy, and how the category of service work in itself is a combination of different categories, some productive, some with little influence in production, and some completely separated from production. By using the Marxist notion of productivity, he is able to show that automation is not possible for every category of what we call services.
But this is not all: Smith takes to heart the comments from Marx in the Machinery and Modern Industry chapter of Capital, to show that automation is not, paradoxically, automatic. When automation happens in one sector of the economy, there is less incentive to automate another sector, since the “supply” of labor power increases and thus wages fall, in some cases even lower than the cost of the machinery that would replace them.
This already happened in Marx’s time. In the aforementioned chapter he shows how in England the number of workers was lower than the number of personal servants. Smith takes this idea to talk about a “servant economy”, an economy that is unproductive in Marx’s term, in which most workers have tasks like teaching, elder care, health services, etc that cannot be automated because they involve a human physical/emotional factor. As Marx predicted, with the advance of machines in production, more and more people would fall into these “unproductive” activities, with lower salaries and worse working conditions.
I have broad agreement with this argument, I think it is elegantly argued, and follows closely Marx’s theory. However, the question is once again What is to be Done? Here Smith has an interesting argument. He says that industrial workers have now lost some power, since the threat of offshoring has made Unions more amenable to agreements with Capital. Instead, there are increasing possibilities in, for example, teachers.
This is because teachers are not only (maybe not even primarily) providers of knowledge, but they in fact provide child care, something that was made very clear during the pandemic. I have to say that this argument is interesting but I do not feel it is entirely accurate. First, there is the problem that offshoring has limits, as we can see in the Chinese case. Although for years there have been threats that China would suffer from offshoring, it seems like the supply chain power of the country has proved resilient, at least for the moment. That means that any negative effect on the Chinese industrial supply chain is felt across the world, and there are no easy replacements. This gives Chinese workers in particular a lot of power.
Second, there are some limits to teachers’ action that are related to societal support. Smith notes how in the US, there teachers’ strikes were supported by parents. And that is a key element. If we see, for example, Argentina, the fighting power of teachers has eroded since successive governments have worked hard to break the potential alliance of teachers and parents. Because the main power of teachers is the refusal of provision of child-care, this has a negative effect on parents, who have to juggle care and work. This means that the alliance between both groups rests on tenuous grounds.
That said, these are just some brief objections that I think wouldn’t be rejected by Smith. The point is more that the political and economical power from workers groups is deeply contextual, depending not only on their “place in the production system” but also in circumstantial conditions. In this, Smith is smart to signal that some groups that have been historically forgotten by some Marxists, owing to a misreading of Marx among other reasons, should be reevaluated as workers that could provide the key for the advance of the socialist cause.
April 22, 2023
Wichtiges Buch mit weitreichenden Implikationen für die Arbeiterbewegung. Leider abstrahiert der Autor, obwohl er John Smith's Buch zum Thema zitiert, vom Imperialismus, so dass dem Buch eine sozialchauvinistische Verzerrung anzurechnen ist. Ausserdem auch ziemlich repetetiv, was wohl daran liegt, dass man den Kern des Arguments locker in ein Paper hätte packen können, aber stattdessen ein Buch daraus gemacht wurde. Trotzdem, wichtiges Buch.
May 12, 2022
Read for work. Fantastic Marxist analysis of how labor productivity stagnation is explained lack of investment in tech these days. Thesis is that automation has uneven if not contradicting effects on different parts of the economy. Really good stuff.
Read
October 21, 2022Fantastic read, helped me square some econ concepts like productivity with my own working experience into a helpful analytical tool. The experience I'd liken it to would be when you first figure out how to properly place the z-shaped tetromino in a tetris block.
September 24, 2024
i don't buy the declining rate of profit/declining investment stuff but i looooove the service work analysis in the latter half. really fantastic. ending with boggs is genius. bad title.
Displaying 1 - 5 of 5 reviews