Economist Paul Mason's blockbuster manifesto Postcapitalism suggests that markets just can't organize products whose major input isn't labor or material, but information, and that means that, for the first time in history, it's conceivable that we can have a society based on abundance.

Mason is one of those fortunate writers who can talk about complex economic subjects and still be engaging, and Postcapitalism is a grabber from page one.

Of course, he's got a lot to work with. Global capitalism has gone weird. Whole generations are unemployed, pension liabilities are skyrocketing, and the decades-old assurance that acting selfishly will make the world better for everyone is a hollow joke. Occupy, Gezi, Arab Spring, Hackney riots, Black Lives Matter, Ferguson, mass surveillance, Euromaidan, Corbyn, Sanders… Where is it all heading? How did we get here?

Mason constructs a careful argument, grounded in the failings and successes of Marxist theory. Specifically, he's interested in the relationship between Marx's Labor Theory of Value to information technology. In the Labor Theory of Value, the true value of goods is the amount of labor it requires, including the labor included in delivering that labor — the hours that went into growing the worker's food and making her clothes and getting her to the factory. Automation produces more goods with less labor, driving the value of goods down, so industry seeks out more complex products that need more labor, and more opportunities to profit.

If you accept this — and Mason has quite a long argument about where the Labor Theory comes from and how it's been tested and what counterarguments have been raised and so on — then it's clear that as goods become more and more made from information, not material or labor, that their prices get very weird and hard to determine. The marginal cost of reproducing an information good — a song, a book, a movie — is zero.

That's not to say that books and music and software cost nothing to make. There's a bunch of fixed costs associated with their production — the labor that I put into writing my books and the labor of the farmers who feed me and the spinners who clothe me while I do (not to mention the factory workers in China who made the laptop that I type on, and the miners in Africa who extract the conflict metals in that laptop). But to arrive at the cost of any individual copy, you have to divide the fixed costs of production by the number of copies that can be made once the book is done: an infinite number of copies. Divide any number by infinity and you get zero.

Unlike machines, information doesn't wear out from use. Software "machines" can theoretically do an infinite amount of work, in an infinite amount of places, for only the cost of electricity and computers to run on. As our goods become more and more conjured from information — as everything from jet engines to pork noodles are — they become more abundant, and the case for using markets to organize our world falls away.

So far, so good. I've made this argument before in fiction and essays, and even mentioned a lot of the same technological systems and phenomena as Mason cites.

Mason makes good and often brilliant arguments about the transformative effect of information technology on market capitalism. They're important arguments, too. But often, they're just not very informed arguments when it comes to technology.

For example, in a section on how the positive externalities from companies like Amazon and Tesco could be socialized, he suggests that user data could be "suitably anonymized" and then made public, which sounds like a plausible idea, but contains one of the most contested technical arguments in policy circles today: namely, whether there is any basis for claiming that large data sets can ever be de-identified in a way that doesn't allow for mass-scale, automated re-identification later.

Mason likewise misses the deep divisions in the open source/free software movement, misidentifies the founder of Creative Commons, and never really comes to grips with the deep debates over the moral versus instrumental cases for free and open licenses. He cites Benkler's Wealth of Networks, but for all his discussions about copyright, never mentions Lessig (let alone Samuelson). And though the casting of products as embodied information is at the core of his argument, he has apparently never heard of spimes, nor read Sterling's Shaping Things.

These are glaring omissions, and they show. Mason is entering a debate that has been conducted for a long time, and which has been defined by contributions from real experts on software, 3D printing, information theory, copyright, crypto, licensing, anthropology, sociology, and a myriad of other disciplines. I disagree with some of those people and agree with others, but the point is, they've covered a lot of turf.

I'm not arguing that Mason hasn't paid his technical dues and therefore does not deserve an opinion. Rather, Mason's argument often rat-holes into easily discredited cul-de-sacs of technological hand-waving and long-settled groaners, and it is to his book's detriment.

Which is a pity, because there is a lot to love in this book. It's audacious and when Mason is arguing his own topic, economics and revolutionary history, it is thorough and crisp as you could hope. There's arguably too much history of Communist revolutionary movements here, seemingly aimed at Old Left skeptics who need to have their concerns about how this accords with historic disputes in Marxist circles.

This is a book that plays to Mason's strengths, but displays some regrettable weaknesses. It's fine to treat software as a theory-object in which it is a machine that never wears out. Just as physicists need to descend from abstract frictionless surfaces and geometric solids and engage with the stubborn materiality of the physical world, Mason's particular argument about infinitely reproducible machines with no marginal labor inputs needs to address the fact that code is a liability, not an asset.

Likewise, it's great to talk about modelling economies in depth, but there is a mathematics of complex systems modelling that is engaged in lively debate about the limits of big data, and it would be nice to hear how Mason's plan takes account of those limits, when (not if) we reach them.

Thomas Piketty's Capital in the 21st Century failed to come to grips with Marx. Mason's Postcapitalism spends too much time on Marx, and not enough on networks. But both books are indispensible reads, with important things to say.

To get a flavor of Mason, try longread in the Guardian, The end of capitalism has begun.

Annoyingly, even though the UK edition has been out for a month, Mason's US publisher has scheduled his book for February 2016. It's worth getting a copy from the UK now.

PostCapitalism: A Guide To Our Future [Paul Mason/FSG]