A digital capitalism Marx might enjoy
Private ownership of capital is the defining feature of most of the world's economies. The conflicts between the owners of capital and the laborers who operate it for them defined two centuries of history. Not for nothing did Karl Marx title his indictment of industrial economies, simply, Capital. Yet the nature of capital changes with time and technology. The world may soon face a new era of conflict between labor and capital, based on a relationship between the two far different from that which animated Marx.
For most of industrial history, capital meant tangible things like looms and furnaces and other machines that you could see and smell and fall into if you were insufficiently cautious. Capitalists spent heavily to outfit their factories, and they put the maximization of these factories’ output above all else. But they also depended on a growing workforce to operate the machines. Capital and labor each sought to prevent the other side from gaining the power to dictate the terms of the relationship—and the distribution of profits generated by it.
Today’s corporate giants rely on different sorts of capital entirely, with very different demands. In their recent book, Capitalism Without Capital, Jonathan Haskel and Stian Westlake describe a 2006 analysis of Microsoft. The company's market valuation at the time was about $250 billion. But its book value was only $70 billion, largely cash and financial instruments, while a mere $3 billion or so could be ascribed to what is usually thought of as capital: plants and equipment. Almost all of Microsoft’s worth was in intangible assets like its intellectual property and brands. Intangibility is most pronounced in tech firms, but it’s important across the economy. A recent analysis found that less than 20 percent of the market value of S&P 500 firms was due to the tangible assets on their balance sheets—a reversal of the ratio that prevailed in the 1970s.
Today most capital, in value terms at least, lives in neurons and silicon rather than on factory floors. The computerization of everything from toothbrushes to pickup trucks means that ever more of a good’s value derives from the software that operates it. The know-how needed to design and build such products (and to manage the complex supply chains that actually produce them) is yet another component of intangible capital. The growing power and appeal of AI stretches the definition of capital still further. Machine-learning programs are an odd form of quasi-labor, trained on data generated by people to do tasks previously done by people. Yet they are owned and controlled by firms in the same way a truck or computer would be.
This evolution fundamentally changes the relationship between labor and capital. While the world of industrial capitalism was shaped by the conflict between the two, there was nonetheless a certain balance of power, since they also needed each other to unlock the riches made possible by technological change. Digital capitalism is different.
On the one hand, as machines grow increasingly autonomous, capital will need fewer workers. In the industrial era, machinery was a substitute for some workers but a complement to many others, such as the tens of millions of relatively low-skilled workers needed to operate factory equipment. Ever more capable AI, in contrast, is very nearly a pure substitute for labor. As it spreads across the economy, labor will lose both leverage within the workplace and the moral claim to a share of the economy’s profits that working provides.
Yet on the other hand, labor is not really becoming less essential, at least not yet. To a great extent, intangible capital is people. Within the elite firms developing and deploying the technologies that are changing the economy, the most valuable corporate capital is the culture—the procedures and norms that shape interactions between highly skilled workers, turning their individual expertise into profitable new ways of doing things. This culture is not like computers or robots; it lives in the heads of workers, who modify it all the time and pass it along to new colleagues.
All the same, labor is being weakened. Even those workers who remain indispensable struggle to capture the returns on the intangible capital to which they contribute. An effective firm culture is a competitive advantage, which cannot easily be replicated by upstart competitors and which workers cannot credibly threaten to take with them if they go. Within the world’s most valuable companies, asymmetric bargaining power allows the returns on this cultural capital to flow mostly to shareholders.
Across the rest of the economy, meanwhile, technology squeezes workers’ bargaining power by giving firms ever more scope to automate or outsource jobs when employees get too fussy or demand pay raises. People whose personal data makes up much of tech firms’ value cannot claim a share of that value either.
This problem will grow more severe over time. The AIs that promise to displace millions of workers are just clever aggregations of countless human actions and communications. Across most of the workforce, capital is learning from labor in order to mimic labor and, eventually, replace labor—all without compensating labor for its enabling role in this process.
Having lost their leverage in the workplace, workers might instead use the ballot box to secure more of the capitalists’ wealth, whether through tax reforms that give fewer breaks to owners and shareholders, and make it cheaper to invest in people, or in the more radical form of a basic income or government--provided make-work. But while such strategies might save people from poverty, it would not recognize workers’ earned right to the economy’s bounty—only the state’s responsibility to provide for those who cannot provide for themselves.
In a new book, Radical Markets, Eric Posner and Glen Weyl describe a very different way to give people control of, and a right to the value of, their contribution to capital. The proposal: treat the data we generate while talking to Alexa or liking things on Facebook as the output of a job of sorts, for which the big tech companies ought to pay us a wage. In other words, treat the data these firms amass as labor, not capital.
In such a world we might, on liking a friend’s photo, be asked by the social network of our choice to provide some contextual data in exchange for payment. Getting paid for our data, Posner and Weyl suggest, could mitigate the harm of mass unemployment, recognize what people contribute to production even if they don’t work at a company, and perhaps give the economy a productivity boost too, since companies would find it easier to obtain high-quality data. Perhaps, they say, the data generators of the world could unite and form a data union, the better to negotiate with big tech companies on fair terms.
But this might all prove too clunky or difficult to organize. Do we really want to spend our days providing metadata to big companies in exchange for micropayments? And would those payments be enough?
Instead, society might settle on a different, collective approach. Data itself could be considered a public resource. The companies that gather data might be required to provide open access to anonymized versions of it (perhaps after the expiration of a short “data patent,” which would reward the company that took the trouble to collect it with a brief period of exclusive use). In exchange for the right to access the data, firms could pay the government an annual royalty, which it might distribute across the population.
Or the government might begin taking ownership in firms itself. Giant sovereign wealth funds might buy shares on behalf of the data-generating public. Dividend payments would enrich the fund, which could in turn pay dividends to the public: the just reward for their contribution to production.
Of course, there is no reason governments can’t do this right now; indeed, some essentially do. Norway, for instance, operates a sovereign wealth fund worth more than $1 trillion, which owns substantial stakes in many Norwegian companies; its returns help fund an extraordinarily generous welfare state. But the case for such a radical approach grows as information accounts for more of the indispensable capital in the economy. A giant piece of mechanical equipment can be used by only one firm at a time, and for only so long before it deteriorates. We have private property rights and free-market competition so that such equipment can find its way to its best use. But the information in our data can be replicated and reused endlessly. The best way to make sure it finds its best use is to allow anyone to access it, under appropriate conditions and in return for fair compensation to society. With new capital comes a new capitalism—perhaps one, finally, that Marx could warm to.
Ryan Avent is a senior editor at The Economist and the author of The Wealth of Humans: Work, Power, and Status in the Twenty-First Century.
Geoffrey Hinton tells us why he’s now scared of the tech he helped build
“I have suddenly switched my views on whether these things are going to be more intelligent than us.”
ChatGPT is going to change education, not destroy it
The narrative around cheating students doesn’t tell the whole story. Meet the teachers who think generative AI could actually make learning better.
Meet the people who use Notion to plan their whole lives
The workplace tool’s appeal extends far beyond organizing work projects. Many users find it’s just as useful for managing their free time.
Learning to code isn’t enough
Historically, learn-to-code efforts have provided opportunities for the few, but new efforts are aiming to be inclusive.
Get the latest updates from
MIT Technology Review
Discover special offers, top stories, upcoming events, and more.