Zuboff’s definition of surveillance capitalism in “The Age of Surveillance Capitalism” commits a category error, a crucial misstep in understanding platform technologies
Shoshana Zuboff’s book “The Age of Surveillance Capitalism” set a new tone in tech analysis since it was published in January of 2019. In just a year, her term, surveillance capitalism, has become a dominant mode of analysis. This year I am going to read the book and then post raw notes and comments on the book. What follows are my thoughts on the first part dealing with the definition.
Zuboff spends little time getting to the meat of the argument. On page 8 under the heading “What is Surveillance Capitalism,” Zuboff describes this new world,
Surveillance capitalism unilaterally claims human experience as free raw material for translations into behavioral data. Although some of these data are applied to product or service improvement, the rest are declared as proprietary behavioral surplus, fed into advanced manufacturing processes known as “machine intelligence” and fabricated into prediction products that anticipate what you will do now, soon, and later. Finally, these prediction products are traded in a new kind of marketplace for behavioral predictions that I call behavioral futures markets. Surveillance capitalists have grown immensely wealthy from these trading operations, for many companies are eager to lay bets on our future behavior.
To summarize surveillance capitalism, which is highlighted by her emphasis, users create behavioral surpluses that are crafted into prediction products and then sold on behavioral futures markets.
By couching surveillance capitalism as a futures market, Zuboff ties herself to all of the baggage of an already existing market institution with a very large literature in economics. At first, I thought that it was a risky move, since futures markets don’t clearly apply to Google or Facebook or Microsoft. But now, after getting further into the book, I tend to think it is a fairly egregious category error.
By way of background, futures markets are typically contrasted to spot markets. Futures markets deal with products or commodities that are delivered in the future and spot markets deal with products that are delivered immediately. The ad markets underlying Google and Facebook are best understood as spot markets, where advertisers bid on clicks and views at the time that they occur. Advertisers might buy a large number of ads all at once that are placed over time, but ad inventory and placement doesn’t magically turn a spot market into a futures market.
If you aren’t convinced these markets are spot markets, consider what it means that “many companies are eager to lay bets on our future behavior.” If I make a bet about the future price of corn in a futures market and it doesn’t work out, then I pay up. I don’t say this lightly. I did very poorly in my graduate financial economics final because I didn’t properly price corn futures. That brings me to my snarky question for Zuboff: What’s the Iron Condor for surveillance capitalism?
By extension, if Facebook and Google were truly selling products that anticipated what people will do in their future, then Clinton’s 2016 campaign should have gotten a massive refund. Indeed, if every ad were a prediction product, then the social media companies would be on the hook when the predictions don’t work out. But these platforms don’t sell prediction products, they sell ads.
By my reckoning, there is only one group trying to work on predictive markets of the kind that Zuboff outlines and it is run by MIT Professor Munther Dahleh. As he explained to MIT Sloan Management Review:
Right now, companies that are getting into the business of trading data are buying and selling data sets, with a fixed price. Buyers don’t know how much financial value the data is going to provide, or if it will provide value at all. If you don’t know what value a data set will provide, I can’t tell you how much you should pay for it.
But we can price the value of a prediction. Getting back to my retail example, if you’re trying to predict your inventory, every improved percentage of prediction translates to a dollar value for you. If I tell you, “This is how many jeans I predict you’re going to sell,” you know exactly how that translates into your bottom line. You know exactly how much an accurate prediction is worth to you, and that determines what you’re willing to bid for it. When you make a fair bid that reflects the value of the data, the market can price the data appropriately.
Indeed, Zuboff could have used the term prediction markets, which is slightly closer to the idea she wants to invoke. Prediction markets are new, having sprung up in the past decade with the release of Intrade and Betfair. But again, these markets are two way bets. Both parties agree to a price on a future event and then settle up once it has occurred. That isn’t happening on tech platforms.
More commonly, analysts unify big tech companies under the concept of multi-sided platforms. The business model has been around for several centuries, but with the advent of the Internet and a massive decrease in computing, it is easier to bring together two or more interdependent user groups. Social media sites and search engines are only two examples of this model, which also includes gaming consoles, credit cards, and shopping malls. For Facebook and Google, the value in these platforms comes in marrying various kinds of content that users demand with an ad market that advertisers and marketers want. Separated, neither side would command much value.
Zuboff also has non-traditional notions of value, claiming that,
Surveillance capitalism’s products and services are not the objects of a value exchange. They do not establish constructive producer-consumer reciprocities…Surveillance capitalism’s actual customers are the enterprises that trade in its markets for future behavior.
Of course, we know that users value “Surveillance capitalism’s products and services” to some extent, otherwise they would do something else other than spending over an hour each day on these platforms. Research confirms that these services are valued. The median user would require \$17,530 to forgo search engines for a year, \$8,414 to stop using email for a year, and \$3,648 to go without digital mapping technology.
The proper term for the problem Zuboff lays out is the externality. Google and Facebook are interacting with advertisers, but in doing so, they impose an external cost on a third group, users. Yet again, there is a deep literature on this subject and its application in multi-sided platforms, which isn’t referenced. Technical papers as well as the advocacy community often start their analysis of big tech platforms with this kind of setup. Consumers are being harmed as a necessary product of the more important commercial interaction, which are selling ads.
Even though I am just a couple of chapters into this book, my initial reaction is one of disappointment. Zuboff is clearly a gifted writer and excellent at turning a phrase, but the analysis makes some very obvious mistakes that are hard to ignore. Upfront, it worries me that Zuboff uses the term futures market, which has a very specific usage in a related field but does little to explain why her vision differs from the entire economics discipline.
Instead of resting surveillance capitalism on a kind of market institution, Zuboff should have relied upon a kind of market logic. Storage is cheap and companies face little downside in collecting and organizing information about its users. If surveillance capitalism were defined that way, it isn’t a far leap to suggest that the current market logic pushes companies into ever expanding data collection that might be harmful to consumers. Yet again, this idea isn’t new.
In the coming weeks I will have more comments on the book, but for now I would suggest other readers take a look at research on markets and multi-sided platforms.