Economies of Scale, Part III: Power, Inequality, and Dependence

Living in the modern world entails existing within an extensive division of labor, one that is ultimately driven by economies of scale. As noted in my previous post, that division of labor allows us to find our niche in society by specializing. We become dependent on a vast array of other individuals, but we also have opportunities for developing specialized skills, interests, and occupations. But scale has many downsides as well that are at the root of many of our society’s problems.

The first and most obvious problem with scale has to do with concentrated power and inequality that comes with it, a problem that has been recognized under the headings of antitrust or competition policy for a century and a half. Large firms pose two kinds of threat: first, economic, whereby control of a particular market allows a firm to charge monopoly rents and to use market power in a host of nefarious ways to disadvantage competitors.  The second in political: in democratic societies, large firms can buy lobbyists, fund think tanks, and wage PR campaigns. 

We have a particular problem with large firms in the United States. One of the consequences of the rise of “neoliberalism” and the Chicago School in the 1980s was the broad critique of antitrust policy made by academics like Robert Bork and George Stigler. They argued that maximization of “consumer welfare” should be the sole criteria for judging the impact of scale, and that most large firms were large because they were more efficient. This was in line with the general tendency within neoliberalism to place economic efficiency above other kinds of social goods, and to look the other way with regard to monopoly’s concentrated political power. At a certain point the efficiency argument begins to fail as large firms spend more time suppressing competition than innovating; the economist Thomas Philippon has presented evidence that prices in the United States have exceeded those in Europe in recent decades due to the European Union’s stricter enforcement of competition laws.

In the digital age, scale has caused several unique problems. Network economies mean that the larger a network grows, the more useful it becomes to its members. This explains the dominance of the internet by a small handful of large platforms like Meta and Alphabet. This is particularly problematic for a democracy because these platforms don’t just sell cars or handbags; they traffic in politics, opinion, and information. While there is a huge cacophony of voices on the internet, the large platforms have the ability to quietly shape political conversations in ways that legacy media companies never could. ___STEADY_PAYWALL___

In 2020 I led a “Stanford Working Group on Platform Scale,” in which we argued that the real problem posed by the large platforms was not trafficking in misinformation and hate speech—most of which was protected by the First Amendment—but rather the ability of the platforms to quietly amplify or suppress political speech. We do not want the government to determine what is or is not acceptable speech, but we also do not want large private for-profit companies to make those determinations either. We likened their power to a loaded gun lying on the table in front of us, where we hoped the individual sitting on the other side would not pick it up and shoot us with it. 

This scenario has been playing out before us since Elon Musk’s takeover of Twitter. Twitter’s former owners did trend left in their content moderation policies, something that irked Musk enormously. It would appear that his purchase of the company was motivated by a desire not to protect free speech, as he argued at the time, but to make the platform friendlier to conservative voices like his own. And indeed, since his purchase, the platform’s content has shifted notably in a right-wing direction. 

We may have differing views of the desirability of this shift, but it should not be up to a single rich individual to be able to shape democratic discourse in this fashion. The solution our working group came up with was something we called “middleware,” where the large platforms would be made to outsource content moderation to a competitive ecosystem of third-party companies that would moderate content based on the choice of individual users. This has been at the core of an amicus brief I participated in contesting new laws in Texas and Florida whose intent was to force the platforms to carry conservative political points of view. 

There is a second set of less well-recognized social problems created by large scale which go under the heading of “winner-take-all” economics, popularized by the economist Robert H. Frank. Frank argues that large globalized markets have concentrated rewards in the hands of a small number of prominent superstars at the top of hierarchies in music, sports, entertainment, and the arts. Taylor Swift, Beyoncé, and Selena Gomez command huge audiences all over the world because technology has allowed their music to be streamed everywhere, crowding out local stars who would have been far more prominent in earlier decades. 

This has made life tough for talented people further down the hierarchy. The daughter of one of my friends was a superb cellist who has had an enormously hard time finding full-time work. She once went to an audition for a chair at the Richmond Symphony—a definitely second- or third-tier orchestra, and found herself competing with several hundred other talented cellists. In a previous age, the Richmond Symphony would have been supported by local people, but today,  why listen to the Richmond Symphony when you can attend a livecast of the New York or London Philharmonics in your living room? 

There are many other downsides to scale but one of particular concern harkens back to Marx’s original critique of Adam Smith’s pin factory. We now live in a Durkheimian organic division of labor, where specialization and mutual interdependence have been carried to such an incredible degree that it becomes vulnerable to disruption and breakdown. We saw this occurring in the early days of the Covid pandemic when supply chains began to freeze up and parents couldn’t get their infant formula. Americans suddenly realized that they were dependent on a potential adversary not just for computer chips but for basic medical protective gear.  

One might argue that the world’s recovery from Covid is evidence of how resilient the world is, but I’m not sure such optimism is justified. There are many future disasters that are entirely plausible that would disrupt everyday life in even greater orders of magnitude, from a major cyberattack on national infrastructure, to a devastating EMP (electro-magnetic pulse) explosion disabling all electronic equipment, to a new and far more deadly pandemic, to a cascading series of environmental catastrophes like the eruption of the Yellowstone supervolcano, to the use of a weapon of mass destruction, whether biological or nuclear. 

Matt Crawford has recently written a book decrying the whole shift towards self-driving cars that are packed with computers and sensors and relieve drivers of the need to know how to drive.  We are raising a whole generation of middle-class kids who can ask ChatGPT to write a book review, but don’t know how to change a spare tire or operate a welding torch. The newer cars will break down much more easily given their complexity, and their operators will be at a loss as to how to respond. Crawford’s previous book Shopcraft as Soulcraft decried the shift from industrial to post-industrial skills, that is, high schools that taught kids web design rather than welding or woodworking. For Crawford, this represented a loss of fundamental autonomy, the ability to control the conditions of your life. It also represents a loss of resilience should those hi-tech systems suddenly stop working.
These possibilities have engendered a whole genre of post-apocalyptic science fiction movies and novels based on the breakdown of our current civilization. Anticipation of such a future has motivated the survivalist movement, which is busy stocking concrete basements with canned food, water, and guns. Worrying about personal vulnerabilities in an anticipated civilizational collapse is definitely not a crazy concern. 

What is harder to know is the degree of breakdown we need to anticipate, and therefore the kinds of survival skills that are best to retain. Many survivalists assume that the apocalypse will set us back to something like Hobbes’ imagined state of nature, the worlds described in Olivia Butler’s Parable of the Sower or Cormack McCarthy’s The Road where the most important skills are knowledge of how to grow food or use guns. But the post-apocalypse may send us back not to 10,000 B.C. but to the year 1900, when we have lots of mechanical devices but no computers or digital controls.  In that world, knowledge of basic metalworking and primitive communications technologies may be much more valuable. Or the post-apocalypse may take us back to 1950, when Matt Crawford’s favored skills like knowing how to drive a stick-shift car will come in handy. 

So we are living in a world where economies of scale continue to shape social relations, and where the shift from mechanical to organic solidarity continues unabated. This shift is driven by the economic benefits of scale; as noted in the first piece in this series, the emergence of artificial intelligence only intensifies this trend. Whether the emerging world will be a sustainable one remains to be seen.

Francis Fukuyama is chairman of the editorial board of American Purpose and Olivier Nomellini Senior Fellow and director of the Ford Dorsey Master’s in International Policy program at Stanford University’s Freeman Spogli Institute for International Studies.

Francis Fukuyama's Economies of Scale series:
Part I: "Economies of Scale, Part I: Artificial Intelligence"
Part II: "Economies of Scale, Part II: The Division of Labor"