NOVEMBER 24, 2014
IN HIS BOOK Average Is Over, economist Tyler Cowen makes a striking prediction: “One day soon we will look back and see that we produced two nations, a fantastically successful nation, working in the technologically dynamic sectors, and everyone else.” He presents this new order, dubbing it hyper-meritocracy, as the inevitable result of technological change: smart machines will create an economy so efficient it will empower the top 10 to 15 percent of earners, allowing them to live “extremely wealthy and fantastically comfortable lives,” while the rest of society sinks. Cowen’s vision is grounded in a rather basic concept: scarcity. In a world where intelligent machines allow us to create more goods at lower prices, the economic benefits will flow towards what is scarce — land and other natural resources, sellable intellectual property, and certain “unique skills” — and away from ordinary workers, especially unskilled laborers.
Many commentators foresee something similar. For example, Larry Summers, secretary of the treasury under Bill Clinton and then chairman of the Council of Economic Advisers for Barack Obama, has said, “Looking to the future, my guess is that the main story connecting capital accumulation and inequality […] will be the devastating consequences of robots, 3-D printing, artificial intelligence, and the like for those who perform routine tasks.” Certainly the facts of the last few decades, which have seen tremendous productivity gains alongside stagnating wages, seem to indicate that the benefits of technology are not evenly shared. But, wrestling Cowen’s ideas, I can’t shake the feeling that broad statements about smart machines obscure another, more important issue — namely the way a particular set of assumptions about technology and society are themselves shaping the hyper-meritocratic future figures like Cowen and Summers anticipate. While far from perfect, Average Is Over articulates these assumptions more completely than any other book I’ve read, making it an important guide not only to our potential unequal future, but today’s digital economy, where a handful of tech giants have managed to capture an extraordinary amount of market share — and profits.
In her book The People’s Platform, Astra Taylor looks at the economics of the digital economy and notes, “As much as networked technology has dismantled and distributed power in more egalitarian ways, it has also extended and obscured power, making it less visible, and arguably, harder to resist.” As a result, the commercial web is a kind of test case for hyper-meritocracy. Google so dominates the search-engine market that about one quarter of all consumer internet traffic in North America passes through the company; Amazon alone accounts for about 10 percent of e-commerce in the United States, and so on. Cowen sees this tendency spreading to other industries — extending the power of certain major figures, while at the same time integrating it deeper into society itself. At times, he underscores this in incredibly stark language, such as when he compares the digital-era economy to a billionaire “riding in a limousine, with open windows, through the streets of Calcutta.” The situation is straightforward: “A lot of beggars will be competing for the attention of that billionaire, and yet the billionaire won’t much need the attention of the beggars […] This in short is what the contemporary world is like, except the billionaire is the broader class of high earners and the beggars are wealthier than in India.” Just as those supplicants are totally dependent on the billionaire’s attention, the great mass of workers is dependent on the attention of the elite. And they have just as little leverage to demand, or even gain, that attention. Cowen expects the limousine to keep getting nicer and the people in the street to become even more desperate.
While Cowen’s metaphor is rooted entirely in the distribution of power, however, his book largely ignores how power itself works. This is a rather significant oversight. Journalist Sarah Jaffe has observed, “Power is as big or bigger a force as technology in shaping the labor landscape today,” noting that in a global economy, capital can “move where and when it wants,” while the weak post-recession recovery has allowed “businesses to squeeze more and more productivity out of the few workers they keep.” While Cowen is persuasive when he says technology reinforces this, I came away from his book convinced that Jaffe’s overall judgment is more correct — existing power discrepancies are what distort the use of technology, rather than technology distorting society. Technology, after all, is simply a tool. It is disappointing that Cowen never questions how that tool is being used.
Some of that is the nature of his book. Average Is Over reads as almost a self-help tome, primarily concerned with preparing a presumably anxious middle-class audience for how to adjust to the coming hyper-meritocratic order. Even the title seems designed to spook these readers — the whole point of the middle class is that it’s composed of average folk. As a result, the book largely eschews data in favor of a more generalized argument. It also means that Cowen undersells systemic issues like education inequality and the lack of access to technology in certain communities, in favor of extolling the potential of information technology and describing the kind of world this super tech might serve.
These systemic issues complicate matters considerably. An astonishing one in three American households lack access to a broadband internet connection, a problem that is particularly pronounced in rural areas. That certainly puts them at an unfair disadvantage. So when Cowen shrugs off structural inequalities, he seems downright obtuse. The problem only gets more pronounced over the course of his discussion. For example, at one point he blithely compares the coming economic order to the Middle Ages, and actually proposes alleviating some of the effects of this medieval economy by setting aside neighborhoods to “build some makeshift structures […] similar to the better dwellings you might find in a Rio de Janeiro favela.” This kind of insouciance about poverty strikes a discordant note, and creates the impression that Cowen is much more ideologically invested in this Brave New World than he is willing to admit. These flaws, however, actually make the book a very useful primer on this (potential) tomorrow. Precisely because he is writing a simplified introduction to hyper-meritocracy — and because he never questions its justness — Cowen offers a very clear picture of the assumptions, ideologies, and worldview that underlay it.
Just this summer, MIT’s David Autor presented a paper at a symposium sponsored by the Kansas City Fed that examines how the growth of intelligent machines has altered the economy. He concludes that, “The interplay between machine and human comparative advantage allows computers to substitute for workers in performing routine, codifiable tasks while amplifying the comparative advantage of workers in supplying problem solving skills, adaptability, and creativity.” This trend is what Cowen zeroes in on. Anticipating Autor, he declares, “As intelligent-analysis machines become more powerful and more commonplace, the most obvious and direct beneficiaries will be the humans that are adept at working with computers and with related devices for communications and information processing.”
To demonstrate what he means by this, Cowen takes a close look at how computers have changed the game of chess. It’s well-known that the best chess programs can now beat even grandmasters. But what interests Cowen are the ways computers have changed the game for human players. For example, “Humans are [now] more likely to understand when it is possible to leave the King in an apparently exposed position […] because of experience playing with computers.” Good chess programs let talented players simulate a wide variety of games and moves, and then see how different strategies play out. The raw computational power of these programs changes the game of chess from a more aesthetic enterprise into one that is driven by information. As a result, “Top grandmasters are more likely than before to experiment with ‘ugly’ moves […] because now they understand that ugly moves are more likely to work out.” Data trumps elegance.
Cowen sees this kind of symbiotic relationship — where a human player uses the chess program as a tool to become better — as most on display in freestyle chess, a version of the game that lets players consult computers, books, or human advisers. The very best freestylers are able to beat both unaided human experts and sophisticated chess programs. Certain people, while lacking the natural talent of a grandmaster, are simply better at figuring out when to advise the computer and when to let the computer advise them — basically they are good at spotting the right moment to add a little bit of human creativity to the ultimately superior computation skills of the chess software. Cowen sees this as the future — technologically adept workers who are able to add value to what intelligent machines are already doing will be able to out-compete both talented humans and machines. The talented person who doesn’t work well with technology — or with the workgroups information technology enables — becomes less important than the person who is just talented enough to help give the machine a helpful nudge. Meanwhile the person who isn’t able to offer that nudge — the less-skilled person whose job can be totally automated — is left behind altogether. The ideal employee is more efficient than the talented loner, the unskilled worker, and the machine. And that’s why she has a lot of economic power.
Technology also makes it far easier to measure an individual’s productivity. As Cowen puts it, “The better the world is at measuring value, the more demanding a lot of career paths will become.” Those that survive this competition gain leverage. As Cowen himself points out, this trend is already apparent in the online realm — where Facebook, Google, and other giants of the commercial web will often purchase start-ups simply to acquire their engineering talent. Another, even more pertinent example that Cowen overlooks is the Big Board, a continuously updating list of the most popular posts on the website Gawker, which the site makes widely available in house. As Gawker’s founder Nick Denton explained to The New Yorker’s Ben McGrath, the Board is designed to “encourage competition” among his staff. This approach is common on the commercial web. In The People’s Platform, Astra Taylor sites an internal memo by AOL as they purchased the online news site Huffington Post that puts things even more precisely: “Editors could keep track of the money made by each article down to the penny.” In Cowen’s hyper-meritocracy, nearly every employer in nearly every industry will be able to do the same thing.
The people who rise to the top of these kinds of ranking systems will mostly be people who know how to use new technologies in highly effective, fairly unique ways. This doesn’t only mean programmers — he sees marketing and management as other fields that benefit from mechanization. Also, as technology makes more and more jobs “team-oriented,” those that work well with others — personable, team-oriented, and “conscientious” employees — become more desirable. But Cowen’s hyper-meritocrats are first and foremost well-educated. For example, he talks about a hypothetical young person “just out of Harvard with, say, a degree in economics,” who becomes a consultant. Cowen is cavalier about this mythical consultant’s actual abilities, saying, “our society does not know what else to do with these people, who are otherwise not very productive.” But he insists “they really are needed” because their “general intelligence” allows them to see problems that more specialized workers do not. Interestingly, though otherwise skeptical of university bureaucracies, he doesn’t doubt the ability of elite schools to skim off the very smartest workers.
It’s not hard to see the elitism that undergirds this system. First, there’s the aforementioned problem that lots of households don’t even have a broadband connection. Common sense would say that when the ability to work well with machines becomes essential to economic success, people who grow up without a decent internet connection are at a distinct disadvantage — forced to catch up in adulthood to people who grew up taking advanced communication technology for granted. This is especially important when combined with an environment that associates “general intelligence” with a particular set of educational credentials. The imbalances between the best private schools and the worst urban and rural public schools are already well-known, but Cowen expects this gap to become even more pronounced. While he extols the virtues of digital education tools, and sees their ultimate triumph as inevitable, he offers no illusions about their effectiveness: “Online K-12 education has a future for many of the same reasons that McDonald’s has been a successful corporation, even though there can be little doubt that a Big Mac is not nearly as good as the finest dish in the best restaurant in town.”
Though Cowen doesn’t see it, the future he lays out seems rife with obvious, intrinsic structural inequalities that will make it very hard for anyone born outside the elite to actually show enough “merit” to rise into it. And when he breezily asserts, “The more that the high earners pull in, the more people will compete to serve them, sometimes for high wages, and sometimes for low wages,” and that, “making high earners feel better in just about every part of their lives will be a major source of job growth in the future […] Better about the world. Better about themselves. Better about what they have achieved,” it becomes hard not to see this as a new form of aristocracy — one where people born with certain advantages are able to leverage them even further than today’s wealthy. Certainly, a smart, capable aristocracy, one theoretically open to talented outsiders, but an aristocracy all the same.
Cowen is careful to note that this system “is not necessarily a good and just way for an economy to run,” but he certainly sees it as a given. Interestingly, he is also keen to emphasize the autonomy of the individual in the hyper-meritocracy. This isn’t itself surprising. But Cowen’s efforts to square the system he anticipates with humanistic ideas about individual agency fall flat. When he defends the possibility of building third-world style slums in the United States, he insists, “No one is being forced to live in these places […] I might prefer to live there if my income was low enough.” Cowen essentially defines choice down to the absence of force. But this is meaningless — after all, no one chooses to live in a slum, unless the alternative is homelessness. Choice only matters when there are real alternatives to pick from. When Cowen compares a hyper-meritocratic society to the Middle Ages, he does so merely to point out that it is possible for a deeply unequal society to remain stable over a long period of time. But the comparison brings to mind another thought instead — that the values that underlie hyper-meritocracy are as un-humanist as those of the Medieval period.
That becomes more evident when contrasting his view of IBM’s Deep Blue — and chess programs more generally — with that of computer scientist/commercial web skeptic Jaron Lanier. Deep Blue, a chess-playing program, became famous in 1997 when it defeated grandmaster Gary Kasparov in a match. Cowen remains enamored of the program — using it as a shorthand for computer intelligence, such as when he muses about the uncertain future of dating algorithms: “It remains to be seen how [a picky dater] will respond when a machine as smart as Deep Blue sends her to a forty-two-year-old in Kansas as a potential mate.” In You Are Not a Gadget, however, Lanier sees Deep Blue as a purely human achievement: “What happened was primarily that a team of computer scientists built a very fast machine and figured out a better way to represent the problem of how to choose the next move in a chess game.” To Lanier, calling Deep Blue smart projects certain human qualities onto a machine, instead of looking at it for what it is: a tool created by humans.
Writing several years before the publication of Average Is Over, Lanier anticipates Cowen’s enthusiasm for an economy built on human-tech collaboration, but sees it as misplaced: “[T]he idea of AI has shifted the psychological projection of adorable qualities from computer programs alone to a different target: computer-plus-crowd constructions.” Lanier’s objection is rather straightforward — when we project intelligence onto a computer, we “have abandoned aspects of the subject at hand in order to remove from consideration whatever the computer is blind to.” This is a problem because it limits both the creativity of the user and the impetus to make the software more responsive: “When people are told that a computer is intelligent, they become prone to changing themselves in order to make the computer appear to work better, instead of demanding that the computer be changed to become more useful.” Cowen is certainly aware of this problem, at least to an extent — he understands that “when the computer plays the opening [of a chess game], it isn’t the computer ‘thinking,’ but rather it is the computer having recorded the best of human knowledge about chess openings to date.” But he also is sanguine about a future where most workers will indeed have to adjust to the needs of a computer: “The worthless worker is one whose cooperation with the machine makes the final outcome worse rather than better. A potentially valuable worker offers the promise of improving on the machine, taken alone.” The tool determines the worker’s value — not the other way around.
From a certain angle, what Average Is Over describes looks more like a system of ethics than one of economics. At one point, Cowen makes a casual observation that really drives this home: “it’s pretty common to hear tenured economics professors at establishment schools espouse the relevance of liberal democratic policies, such as the social safety net. These same individuals, if asked to explain their choice of academic hires […] often respond in harshly meritocratic terms.” Regardless of what you think the proper role of safety-net programs like unemployment insurance or Social Security, it’s pretty obvious that Cowen is making a rather dramatic leap here. After all, the point of the safety net is that it isn’t based on merit — it’s based on need. It’s quite possible to support, say, a program to provide free medical care to the indigent (let’s call it Medicaid), and at the same time think that it’s a good idea to make sure that, for example, someone graduates medical school before they’re allowed to perform surgery. But this kind of distinction doesn’t really exist in a hyper-meritocratic worldview. People are valued for their economic use and nothing more, and no other claim is really acknowledged. Essentially it elevates the market economy from a tool into an ethical system, the same way it elevates the computer from a tool to a way of sorting out who is “worthless” and who is not. Far from ideologically neutral, this worldview is exactly what makes hyper-meritocracy possible.
When you look closely, hyper-meritocratic ethics pervade society right now. One of the reasons people continue to respond to the work of Ayn Rand is precisely because she invests market outcomes with a certain moral force. And it is certainly natural to want to push back against this idea using the same humanist language Lanier does when he rejects those who would have humans redefine themselves to better suit digital tools, rather than redefining those tools to better suit their human users. But as Average Is Over shows, the language of individual agency and human dignity can easily be appropriated to defend inhumane outcomes — again, Cowen’s hyper-meritocrats aren’t physically compelling anyone to live in those favelas. Further complicating the debate, private companies often get behind projects that will benefit ordinary citizens — Google’s efforts to help provide public broadband service in cities like Kansas City will certainly help alleviate some of the digital divide. But platitudes about “choice” and the occasional corporate good deed don’t have much impact on the intellectual assumptions that shape how tools like technology and the market are used — they only make it easier for those in favor of the status quo to change the subject. The challenge for those who reject the hyper-meritocratic system of ethics is to find a point of view capable of offering a genuinely alternative measure of value.
Again, the most obvious place to look would be the commercial web. In The People’s Platform, Astra Taylor floats a new way of talking and thinking about culture online: cultural democracy. Taylor’s book has a much narrower range than Cowen’s — this is for the better, both because it allows her to utilize her research in a much more explicit way, and because her book is significantly more focused — but her observations about the commercial web extrapolate well in the wider hyper-meritocracy of the digital era. Like Cowen, she sees how the efficiencies of the digital age can magnify inequality, noted that “The Internet does not close the distance between hits and flops, stars and the rest of us, but rather magnifies the gap, eroding the middle space between the very popular and the virtually unknown.”
Unlike, Cowen, however, Taylor interrogates this situation. When faced with the rise of “productivity gains that fail to improve stagnant wages,” she bluntly declares, “Our society’s increasing dependence on free labor — online and off — is immoral in this light.” And she focuses a great deal of attention on the underlying inequalities in Internet access, education, and social power that Cowen shrugs off. What’s most important about Taylor’s book, however, is that she realizes the current frame of how we talk about these issues — with its emphasis on technology instead of political power and its automatic acceptance of market outcomes — actually makes it harder to counter the ideologies that make hyper-meritocracy possible. To that end, she proposed an alternative: cultural democracy. This cultural democracy doesn’t pretend that “talent and motivation are evenly distributed.” Instead it focuses on fostering “contact between conflicting perspectives.” She sums it up thusly: “Cultural democracy means that a diversity of voices and viewpoints is expressed and accessible, that visibility and notoriety should not be the consequence of cumulative advantage alone, and that influence within the cultural field is achieved by a variety of factors, not simply ceded to those who can afford to pay to be seen and heard.”
Taylor’s ideas suggest an ethical vision commensurate with hyper-meritocracy — namely hyper-democracy. Democracy and meritocracy have long been in tension, of course, and Taylor’s ideas dovetail with a larger radical democracy movement. To pick on recent example, David Graeber’s The Democracy Project, a political manifesto inspired by Occupy movement, proposes “the gradual creation of a culture of democracy,” one that like Taylor’s wants to foster contact between conflicting points of view in an effort to move towards consensus. Faced with political ideas that work abstractly but fail in the real world, Graeber moves away from abstract “reason” toward a new standard: “reasonableness.” As he explains, “Reasonableness implies a much more sophisticated ability to achieve a balance between perspectives, values, and imperatives, none of which, usually, could possibly be reduced to mathematical formulae.” Cowen’s book brings this idea to mind for a simple reason: a world where a great mass of people are seen as “worthless” simply because they do not have the acumen or experience to use a particular set of tools is not reasonable. And if a particular set of assumptions about market outcomes and merit lead us to elevate a small minority while writing off the great mass of society, then we ought to rethink them.
 Between 1973 and 2011, worker productivity surged by an astonishing 80 percent, while wages only grew by approximately only 10 percent.