NE #2: Cognitive Skills and Affluence
Some time ago I wrote about the effect of technology on how humans communicate. I made the case that these changes are ill-perceived but are so transformational that we need to rethink the terminology for the deep structure of human society. I called our era that of the “networked episteme”. I will expand on this thesis in a series of essays co-authored with Amasia Fellow Mary Marsh.
The touchstone for this work can be found in a quote by Michel Foucault: “History serves to show how that-which-is has not always been; i.e., that the things that seem most evident to us are always formed in the confluence of encounters and chances, during the course of a precarious and fragile history… since these things have been made, they can be unmade.”
What I am trying to get to, as a layperson trying to make sense of things, is a “history of the present” — how did we get to what seem like immutable truths but are in fact products of highly contingent “encounters and chances”?
The Question: How Did Cognitive Skills Become Associated With Affluence?
We can observe, and take for granted, that we live in an era in which cognitive skills are enormously prized. At one end, we find computer science classes for preschoolers. What lies at the other end is indubitable proof that cognitive skills are now the primary driver of affluence.
We can most easily see this in the industries of American billionaires, which are dominated by two fields, technology and finance, that privilege these skills. And the fields are closely related — as we describe below, the “digitization of capital” has completely transformed how finance works. In our era, it is not labor that has been alienated — it is capital.
This was not foreordained. How did we get to the point that a man (for they are mostly men) sitting behind a screen on a high floor in an office building in Manhattan can make, in comfort, millions, perhaps billions, more than the laborer who, at risk to life and limb, helped build that office building? And what lies ahead in the networked episteme?
The Industrial Revolution was the trigger for a dramatic expansion in information and the need to measure and act on new streams of knowledge. Digital technologies have greatly expanded our capacity to store, manipulate, and communicate information. The analysis of this information — which demands cognitive skills — has become the next frontier for economic productivity.
The value placed on these relatively scarce cognitive skills, and the concentration of ownership that follows from the invention of the joint stock company (and its variants, such as general partnerships), has exponentially multiplied the affluence of those who are trained in these areas.
In the networked episteme, absent tectonic shifts in how the world works, these trends will continue and likely accelerate.
What Counts As Knowledge Work?
We should start with a definition. In a time in which nearly all workers process and act on information, the delineation between knowledge work and other work can be hard to draw. For example, a construction foreman today may be responsible for complex documentation and management. Is she a knowledge worker? How about a postal worker? While not always creating new knowledge, he may manage large quantities of important logistical information.
The economist Fritz Machlup, who is often credited as the first to identify the rise of the knowledge economy in his 1962 book The Production and Distribution of Knowledge in the United States, decided only to include those producing new knowledge, while Marc Porat, another important theorist on the knowledge economy, used a more expansive definition that included any work whose output was information. We will use an academic definition descended from Porat’s, while keeping in mind the internal nuances of this broader category.
“Information work occurs when the worker’s main task involves information processing or manipulation in any form, such as information production, recycling, or maintenance. The consequence of information work is information, whether in the form of new knowledge or repackaging existing forms.”
- Jorge Schement and Terry Curtis, Tendencies and Tensions of the Information Age
The Evolution of Knowledge Work
Although knowledge work is often treated as a recent phenomenon, in the words of the distinguished British sociologist Anthony Giddens, “all states have been ‘information societies.’’ All societies require communication and administration, which means all societies have someone engaged in knowledge work. Both the Han and Roman empires conducted censuses of their subjects, for example, and the magnificent Library of Alexandria was managed by a series of prominent librarians.
The limits of transportation played a key role in keeping this knowledge infrastructure small. Political economist Harold Innis identified two kinds of text in this episteme of text– those that emphasize time and those that emphasize space. Those that emphasize time, like stone tablets, would be heavy to transport. Paper, then, was crucial to the development of knowledge, as were centralized entities that could organize the safe passage and central storage of this knowledge.
The early modern period saw the rise of just such entities. The Uffizi Gallery was once the bureaucratic center of Florence. Philip II and Louis XIV grew significant bureaucracies that collected information on their European subjects as well as their burgeoning territories in the New World. The church was just as relevant, if not more, as a producer of knowledge. It had been monks and nuns who shepherded the wisdom of the ancients through the Middle Ages, and the church had a robust network of information through its parish system, which collected information on churchgoers and reported back to the papal authority.
Still, knowledge work was not a “growth sector” in the pre-modern period. Knowledge collection was, for the most part, either a religious imperative, a project of the state, or a hobby for the enormously wealthy. An infinitesimal proportion of the population was doing what we would consider knowledge work, with most informational tasks integrated into other work.
Communication was limited by the speed of transportation — as late as the eighteenth century, a letter could take two months to cross the Atlantic Ocean. Networks of correspondence on which the spread of knowledge relied could be highly volatile. When Marin Mersenne, a major facilitator of one such network, died in 1648, intellectual production was set back for several years.
The advent of mercantilism created some companies with the infrastructure to support knowledge work. For example, the Dutch East India Company received regular reports from their merchant explorers, and the merchants of Venice supported a burgeoning publishing industry — but information was not yet a true commodity. The limits to communication also meant that most merchants with business in distant places employed trustworthy family members rather than rely on the accountability of independent agents.
The Industrial Revolution triggered a dramatic change.
The Impact of Industry
At the beginning of the 18th century, about 0.2% of the American workforce had a job in the information sector. In a country of 9.6 million, the Bank of the United States was run by just three people, and the entire United States government employed around 600. By 1960, the information sector would employ more Americans than any other sector of the economy. How did such a dramatic transformation occur?
The answer lies in the innovations of the Industrial Revolution, which saw the development of both the need for knowledge workers and the means to make their jobs possible. In Division of Labor in Society, Emile Durkheim describes the expansion of markets thanks to the twin innovations of manufacturing, which dramatically increased the number of goods produced, and of transportation, which dramatically increased the potential range of those goods. According to Durkheim, “The producer can no longer embrace the market in a glance, or even in thought. He can no longer see the limits, since it is so to speak, limitless.”
Economic historian James Beniger calls this “the crisis of control”—in order to manage such a market, businesses could no longer rely on one or two people to account for all the complexity of a dramatically expanded business. Tracking individual shipments of corn and grain traveling around the country became increasingly difficult.
If the development of inexpensive paper and printing made it easier to transmit knowledge in earlier periods, the development of a robust system of roads, railways, and canals through which that information could travel was just as significant to knowledge management in the 19th century. With the invention of the telegraph and the telephone, what geographers call ‘the friction of distance’ was eroded (though not completely shattered – communication technologies were still unreliable and extraordinarily expensive).
But while these systems made communication easier, they too were complex systems that needed coordination and management. Fears of colliding trains were not unfounded, and railways operated under up to eighty different times through the 1870s. It appeared as though there may be a limit to the market after all— the organizational capabilities of the humans who ran it.
The Explosion of Knowledge Work
Here was the real opportunity for the expansion of knowledge work. The pioneers of the knowledge economy recognized that the newfound complexity of industry would require coordination, analysis, and communication.
The Erie Railroad superintendent, Daniel MacCallum, developed a robust system of control that relied on a hierarchy of workers who would collect and communicate information. Frederick Winslow Taylor’s theory of scientific management necessitated a cadre of analysts to streamline workflows. Alfred P. Sloan’s famous ‘organization study’ and restructuring of General Motors decentralized the daily management of the corporation, opening up positions for junior managers to develop their critical and administrative capabilities.
The success of these new corporate structures inspired other companies to follow suit. Between 1900 and 1930, the number of clerical and professional workers tripled in the United States, and by 1960, the information sector employed 42% of the American workforce. And unlike previous periods, bureaucratic knowledge work was not primarily a government domain—in 1930, only 7.8% of knowledge workers were employed by the government, and even in 1960, after massive growth in civil service during World War II and the start of the Cold War, the government employed only 14.1% of knowledge workers.
Technologies like typewriters, filing systems, and other organizational technologies helped keep this information organized, but clerical workers were crucial for the wrangling and understanding of this data. Between 1940 and 1980, as immense gains in productivity thanks to technology shrunk the number of industrial laborers, clerical and professional positions grew by about 2.5% per decade.
The Evolution of the Knowledge Worker
Automation is the bogeyman of the factory worker—take one look at the discourse around the Rust Belt and this is evident. But automation also has a significant role to play for the evolution of the knowledge worker. In her 1988 book, In the Age of the Smart Machine, Shoshana Zuboff writes, “As information technology is used to reproduce, extend, and improve upon the process of substituting machines for human agency, it simultaneously accomplishes something quite different. The devices that automate by translating information into action also register data about those automated activities, thus generating new streams of information.”
Informating and automating work hand and hand—whereas clerical workers of the past would have spent time entering and organizing data, the information generated by new technologies can be organized by automated databases, making some functions of the clerical worker obsolete. Instead, the knowledge worker’s value lies in what Zuboff calls “intellective skill”—the ability to think abstractly and analyze the data that a computer system has collected and organized for you.
In 1995, Peter Drucker, the father of modern management theory and coiner of the term ‘knowledge work,’ wrote, “Increasingly, the true investment in the knowledge society is not in machines or tools, but in the knowledge of the knowledge worker. With that knowledge the machines, no matter how advanced and sophisticated, are unproductive.”
With the vast amount of knowledge necessary for the modern corporation, no one person could know everything—success requires a team of specialized workers whose each offered a unique and invaluable bank of skills gained through time spent on education.
This value on intellective skill has only increased over the last twenty years, as knowledge work has begun to bifurcate along the definitional fracture line that we began with. While the job of a clerk of the nineteenth and early twentieth century included both routine and non-routine tasks and was a stepping stone to management, through the second half of the twentieth century, routine and non-routine work began to be sifted apart, with routine occupations falling behind non-routine ones. Despite those equal levels of growth between 1940 and 1980, from 1980 to 2010, professional and technical positions continued to grow at a rate of 2.5%, while clerical positions began to shrink (-1.5%).
The Knowledge Economy and Education
During this period, education became a vital signal of one’s intellective skill and therefore value to a company. A college degree is part of the new American Dream and not without cause— over a lifetime, the average bachelor’s degree holder will make $2.3 million, 84% more than an average high school graduate, with the gap growing for more advanced degrees. All over the internet, one can find credentials in human resources, project management, and software that claim to make recipients more competitive in the labor market.
The importance of education in the knowledge economy raises an important topic of societal discourse—does everyone have equal opportunity for education, and therefore, success in a knowledge economy?
For part of the 21st century in the United States, this did seem to be somewhat true. The 1944 GI Bill of Rights, which Drucker considered one of the most important events of the century, opened up education to nearly eight million veterans, many of whom helped usher in the so-called Digital Revolution. The 1958 National Defense Education Act further expanded education in response to the launch of Sputnik.
But starting in the 1970s, as technological progress continued at a breakneck pace, the education required to fulfill the demand for knowledge workers in the top level of the bifurcated market became less common. According to economists Lawrence Katz and Claudia Goldin, American educational attainment has slowed in comparison to other countries, thanks in part to lower-quality K-12 education and increasing costs of college.
This crisis of supply only raises the premium on ‘intellective skill.’ It also creates a system that reinforces acquired cognitive ability—those with the resources to attain the education necessary for non-routine knowledge work are able to make more, and subsequently provide the same opportunities for their kids.
“Software Is Eating The World”...
In 2011, Marc Andreesen published an editorial in The Wall Street Journal titled “Why Software Is Eating The World,” in which he wrote, “Six decades into the computer revolution… all of the technology required to transform industries through software finally works and can be widely delivered at global scale.”
Consider a typical consumer in New York City. They might wake up in the morning to an alarm set on an Alexa. They check Slack, then order a coffee on the Starbucks app. They call an Uber to get to work, and on the way there, use Apple News to get up to speed on world events. At work, they use Microsoft Office and Dropbox, Gmail and Zoom. They adjust the temperature in their apartment from the office using their Nest thermostat. They find a place for happy hour using Yelp. When they get home, they have groceries delivered using Instacart and wind down with Netflix.
There isn’t a moment of this person’s day that doesn’t involve software.
In today’s market, value is not created by producing a tangible good, but rather by leveraging information to connect supply to consumer demand — to expand the network of possibilities for customers. The value in companies like Amazon, Airbnb, and Uber is not in the products they advertise — for the most part, they don’t own those products or directly employ the workers who provide services on their platform. Instead, what Amazon, Airbnb, and Uber can provide is the manipulation of massive amounts of information.
The tools of the networked episteme have made it possible to collect, store, and transmit massive amounts of information about consumers and markets. Our smartphones can transmit hundreds of data points to computer servers thousands of miles away in just milliseconds. The most valuable companies today use these tools to direct the user to what they want – whether that’s more information (in the case of Google or Facebook) or a good or service (in the case of Amazon).
And as more and more companies develop software, they, in turn, rely on other specialized software companies to fulfill the individual functions they need. For example, Airbnb employs software from Twilio to messaging customers, software from Braintree for managing payments, and software from Amazon Web Services (AWS) to store data and handle traffic to its website. The outsourcing of these specialized services to companies who can scale their operations more efficiently creates a massive market for software that the average consumer never sees. Consider this — In the fourth quarter of 2019, almost 70% of Amazon’s operating profit came from AWS alone.
...And Driving Affluence
With this vast expansion of the prevalence of software, there are vast fortunes to be made in technology. But the implications of the networked episteme also mean that these fortunes are likely to go to a select few.
In 1981, Sherwin Rosen theorized the economics of what he called “Superstars” — those at the top of their field who, thanks to technological innovations, can capture a larger market and thus dominate the field. In his pre-Internet paper, Rosen used the examples from comedy and classical music to demonstrate the effects of this phenomenon. But recently, economists have found this theory relevant to today’s technology-reliant corporate world.
In the networked episteme, where there is essentially no limit to a cloud-based product, there is no need for a third- or fourth-best product — the best product can serve everyone. This means that beyond an initial period of intense competition to be the best, the rewards of a particular market are likely to go to just one company and its executives — think of the irrelevance of Bing and MySpace, the lack of competition for LinkedIn or Yelp. (One exception to this might be Netflix, due to competition from powerful media behemoths like Disney who have the bandwidth to withdraw their goods and services from Netflix, thus undermining Netflix’s ability to fill viewer demand).
In these winner-takes-all environments, there is a further concentration that follows from the prevalence of the joint stock corporation (and its many variations including general partnerships). It is not just that the winning corporations become economically valuable at levels that were inconceivable not long ago; their ownership is concentrated in relatively few hands. A few very, very affluent hands that are attached to brains that have been trained in relatively scarce cognitive skills.
The Digitization of Capital
In the 1960s, the creation of credit cards allowed people to buy goods without withdrawing money from a bank. In the 1970s, the establishment of the fully electronic NASDAQ made trading transactions possible from anywhere and with little human intervention. The rise of the ATM in the ‘80s made it possible to move money without a teller, and online banking, first introduced by Wells Fargo in 1995, makes it possible to move money without an ATM.
And it's not just in the United States that digitization has been immensely impactful. Thanks to the growth of services like AliPay and WeChat Pay in China, three out of four smartphone users made a mobile payment in 2017, moving trillions of dollars without a single bank note changing hands.
As a McKinsey report recently put it, “As data increasingly become the “raw material” for success, financial-services providers need to behave and function like tech firms.” Financial service providers have already started this transformation — nearly a third of Goldman Sachs’ employees are engineers, a higher total than at Google or Facebook. While quantitative analysts, or quants, have in the past been “second class citizens” on Wall Street, their increasing value to financial service providers has begun to elevate their status.
The growing influence of the financial sector in the economy means that the actions of investors are increasingly important. Today, close to 40% of the one percent’s income comes from interest, dividends, and capital gains, vast financial assets that digitization makes it easier than ever to access. According to a study by McKinsey in 2016, 90% of client interactions in financial services are conducted online or via mobile. This means that any moment, millions or even billions of capital can be shuffled around from a skyscraper in Hong Kong, a mansion in Westchester, or a beach in Tahiti.
As I said in the preamble, what I am trying to do is to trace a history of the present for certain elements of modern life. I am not opining here on what I think is right or fair — I am trying to figure out what is, and how we got there. Moral judgments are left to the reader.
Before the Industrial Revolution, innovation was focused on having enough - subsistence. The capacities developed by machines during the Industrial Revolution forever shifted that focus — instead, innovation today is focused on how to manage the abundance we’ve been able to create. Therefore the most crucial resources available in today’s market are information and the skills to wield it.
The networked episteme has seen the manipulation of digitized information become the main driver of affluence in our society. Knowledge work in this era is like much else in the networked episteme—we can resist, acquiesce, or embrace its imperatives.
In the case of workers, their value has increasingly become tied to the knowledge they can produce out of the flood of information available, creating a bifurcation between those who produce knowledge valuable to economic production and those who don’t.
For companies, technology provides the tools by which information can be made valuable, skyrocketing the most successful commanders of knowledge to immense success and resulting in the seepage of software into all sectors.
For the owners of capital, the successful deployment of information allows for immense power to influence the economy. As the relationship between information and power becomes more far-reaching, the ability to employ information becomes the most valuable asset in the modern global economy.