Two weeks ago I collected the major recent anti-trust/competition lawsuits, by regulators and competitors alike, filed against U.S. big technology companies. My point was that, after a long fallow period where these giants received the benefit of the doubt for their successful competitive practices, the public trust has seemed to turn, supporting lawsuits on a wide variety of theories.
Although I wrote to mark the mid-beginning of a trend likely to continue for decades, my article was premature, as a day after our publication Google was sued in an anti-trust action by 38 states. This lawsuit is the first action where I have seen the term “attention economy” stated, defined, and used as the basis for claims. The states use the metaphor of data being a resource, like oil, that that can be captured and refined into something worth selling.
The states claim that Google “uses its gargantuan collection of data to strengthen barriers of expansion and entry, which blunts and burdens firms that threaten its search-related monopolies (including general search services, general search text advertising, and general search advertising).” Setting aside the fact that Google has a significant direct competitor in Microsoft – a company powerful enough to be the subject of its own set of anti-trust suits by regulators and competitors in the past couple of decades – the claims are similar at their core to the anti-trust cases made against AT&T starting in 1974. Google has built an enormous resource so valuable that everyone uses it – like the telephone network fifty years ago – and they are leveraging this resource to 1) enter other fields as a leader, and 2) keep competitors out of their own revenue streams.
There is much to unpack in this complaint and I intend to do so in a later post. Here, as we career toward the blessed end of our annus horribilis (and we hope, not into another), I want to revisit the metaphorical concepts underlying many of these lawsuits. What are data, really, as a legal concept?
First, we need to parse the term. What we call data is history – a description of what happened and who it happened to – and nobody owns history. Of course, only limited aspects of history are recorded for posterity, but the information captured in the modern world is growing exponentially with cameras and IoT devices at every bank and intersection. Fading memories can reduce the impact of history, but computers can keep their historic information for as long as their owners like.
The classification of information at the base of this and many other lawsuits includes two types of data: transactional data and descriptive data. The combination of the two is especially valuable. It helps to know that 100 people bought left-handed baseball gloves, but it can be much more valuable to know that Tommy bought a left-handed baseball glove.
I am using transactional data in its broadest interpretation right now, captured information about every move made in our world. I’m talking about any activity that can be noted and recorded. This includes online searches, browsing to particular websites, remaining at an internet page for ten minutes – or leaving within seconds, watching videos, requesting videos and not watching them, browsing books or cooking utensils, translating phrases. It includes attending church services, riding the bus, walking in the park, visiting friends, and learning to juggle. And of course, it includes financial transactions, both online and off, where you purchase diapers or stay in a hotel room.
Descriptive data is simply information that can help identify you, which can be as simple as a name, address, or email. But for sophisticated analysts like Google, two or three items of information like your birth date, your gender, or even particular search terms may be enough, in conjunction, to identify you. This is why legislators have such a difficult time defining “identifiable” information. Lists of name, address, and social security number work well for laws concerned about restricting identity theft because this limited data is what the thieves need. However, for laws restricting business use of personal data like the GDPR or CCPA, broad – in those two cases impossibly broad – definitions of personally identifiable information recognize that companies can identify a person from aggregations of data that legislators can’t predict ahead of time.
The concepts are not mutually exclusive, as transactional data can be descriptive – regular purchases of feline treats, food, and litter can describe a person as a cat owner – and descriptive data can have clear transactional implications – if we know where you live and work we are likely to know where you order coffee or buy groceries. But it helps to understand the differences between the two types of data if you are considering the legal implications of data ownership and use.
As a general rule, U.S. law does not recognize ownership of data. Neither transactional information nor descriptive information is copyrightable subject matter. There is a line of cases that protects the economic value of certain “hot news” transactional information like the play-by-play call of baseball games for the people who invested in creating those games in the first place, but only for a very limited time, maybe as short as a few minutes, and then the data is available to everyone.
So, no matter what you would like to believe, you don’t own data that describes you or data created by your own actions. It is not possible to own this information. So, if this thing (information) that is no one’s property has value, who gets to exploit its value? As stated immediately above, not the person described or the person whose actions created the data. While the EU protects such information from certain kinds of exploitation and claims that people have a human right to keep certain parts of this information private, no one has seriously offered a regime where you could make money by selling your own data.
Why not? In part, because no one has recognized that you might have an economic interest in data about you or your life, and in part because recognizing and accounting to you for the use of the data would be difficult, and would involve policy decisions we haven’t seriously debated yet. Individuals would need to push Google and others to provide credits for using our data, and the information giants have no incentive to do so. It has been suggested that data subjects should form bargaining collectives to fight for the value of their data, but I haven’t seen any data unionization gain traction. The government would need to step in to make this idea gain serious traction. The market is unlikely to provide us economic management of our own descriptive or transactional data.
Google doesn’t own it either. But Google holds lots of it and can provide transactional data in a timely fashion. (That’s another issue about transactional data. It loses economic value quickly – if I know someone wants to buy a book now, I can sell it now. If I know someone wanted to buy the book last year, that information has different, and likely lesser, value to me.) The new lawsuit compares this data to oil. I don’t agree. I would argue that, if Google’s data is an economically viable resource, the kind of data used by Google is more like a crop that is harvested and milled into something valuable. Google doesn’t pick its data out of the ground or the air, instead it creates and cultivates a place – its search engine – for transactions (searches) to be initiated by people, collecting the descriptive results of the transactions Google facilitated. Placing a camera at an intersection and collecting information about passing pedestrians is more like drilling for oil – you take whatever you find. Google has cultivated an entire ecosystem where people express their needs and desires, and they harvest the information that is expressed there.
So does the fact that Google has created a place and method for people to voluntarily express their information mean that Google has more of a right to that data than anyone else does? Economically and legally, both oil and wheat are commodities that can be sold by whoever holds them and sold first by the person who can collect them. The court will need to decide. The anti-trust laws can punish Google for the way it wields its market power, depending on how that power is defined. But the legal and economic thinking about how data functions in our society can change the way we live our lives, and who gets a financial benefit for the things we do.