The most important charge against the Tech Giants is that they are creating kill zones: no one wants to create new online functionality that Google or Facebook in particular might easily copy. Whether kill zones exist remains a subject of debate, but that hasn’t stopped antimonopolists from wanting to respond anyway by breaking up the Tech Giants.
But there’s a treatment that is more likely to cure the disease without killing the patient—one which has the added benefit of helping us determine whether there actually is a disease to begin with.
But first, let’s consider why kill zones would be a problem if they really do exist.
If kill zones exist, then startups won’t introduce innovative social media functionality into the market because they will know that if Google or Facebook copy it, consumers will prefer the Google or Facebook versions—because those versions will be integrated seamlessly into the rest of the functionality those companies offer—and so the startups will lose out in the market. But because the startups won’t introduce the new functionality, Google and Facebook will feel no need to introduce it either, and so the functionality never makes it into our online lives.
The saga of Vine, Twitter’s erstwhile video sharing service, provides a nice illustration of the kill zone argument. Vine pioneered the video sharing format, but just as it looked poised to take off, Facebook introduced essentially the same functionality in Instagram, and because Instagram was already much bigger, social media users found it easier to embrace the format in Instagram than by migrating over to Vine. The lesson of that episode for someone with a bright idea for the next big thing in social media is: don’t waste your time. So long as it’s something that the Tech Giants can copy, they will, and they’ll win.
To be sure, TikTok shows us that innovation is still commercially viable in social media, the Tech Giants notwithstanding. But the key to TikTok’s success is the algorithms it uses to target videos to users, and those are not easy for the Tech Giants to copy. That means that major, irreproducible innovation is still possible in this space.
But not all good ideas that make our lives better are necessarily major, irreproducible technological steps. If there hadn’t been a Vine, video sharing might have taken a lot longer to invade social media, and that would have been a loss to users, many of whom love the format.
So what to do about kill zones?
I am often struck in reading Nick Lane’s excellent books on biochemistry (Power, Sex, Suicide, and The Vital Question are my favorites) by how much more careful biochemists seem to be in diagnosing problems and sussing out solutions than we are in antitrust. All the more struck, in fact, because biochemistry is a more mechanistic, indeed, easy, field than is antitrust.
In biochemistry, the basic repertoire of behaviors of the smallest units of analysis—molecules—is known with absolute certainty, thanks to the laws of chemistry and quantum physics. If molecule A hits molecule B, we know exactly what will happen. And hypotheses in biochemistry are often testable: you can find a thousand or ten thousand living human bodies in which to observe the biochemical behaviors that interest you.
By contrast, in antitrust, the basic repertoire of behaviors of the smallest units of analysis—human beings—is almost infinite and the subject of perennial debate. We like to assume rational, profit maximizing behavior, but we know, thanks to decades of behavioral economics, that actual behavior is much, much more complex and varied. And hypotheses in antitrust are almost impossible to test on the scales required to produce real knowledge: where do you find ten thousand similar markets to deconcentrate in order to determine whether breakup actually works?
You would expect, then, that antitrusters would be even more careful in diagnosing problems and sussing out solutions than are biochemists. But instead the sheer complexity of the problem seems to impel us in the other direction. It tempts us to boil away the complexity and then find clarity in a residue that bears little resemblance to actual markets. (For the record, I am as guilty of doing this as the next antitrust scholar.)
We should take a page from biochemistry and recognize that while kill zones sound plausible, plausibility is not reality, and so any solution to the kill zone problem must include, as a precondition, some attempt to determine whether there really is a problem. We need to diagnose.
Simply breaking up the Tech Giants doesn’t do that. It’s a bit like the antioxidant craze Nick Lane critiques in his books. It was certainly plausible that free radicals, which do destroy cells, might be the cause of disease, and so molecules that neutralize the free radicals—antioxidants—would be conducive to health. But when biochemists looked closer, they discovered—at least according to Lane—that free radicals are an integral part of the process by which cells regulate the amount of energy that they generate. They’re not all bad, in other words, and so mitigating the harm they do is not as simple as just getting rid of them through the use of antioxidants.
Similarly, a closer look at Google and Facebook’s behavior might reveal something more complex at work than the mere crushing of competitors, and smashing these big companies might, then, not be the solution.
One way to try to determine whether kill zones really are a problem, and, as an added bonus, potentially to treat any problem that does exist at the same time, would be to provide something like intellectual property protection for social media innovations.
Let’s look first at how this works as treatment. By giving startups the right to charge Google and Facebook for any new social media functionality that Google and Facebook appropriate from them, an intellectual property regime would restore the incentive for startups to enter and innovate, regardless whether Google and Facebook use their technology or not. If the new technology is successful, either the startup profits by succeeding independently in the market or by generating licensing income from Google and Facebook if those companies choose to license the technology. The heart of the kill zone fear is that Google and Facebook can take ideas, and the markets that go with them, without paying compensation: this solves that problem.
To return to our example, if Vine had had a right to demand compensation for Instagram’s embrace of video sharing, then Vine might not have become a business failure after all, and the lesson of Vine for social media entrepreneurs today would be that innovation does pay after all, the power of the Tech Giants notwithstanding.
Indeed, intellectual property rights in social media methods would not only encourage innovation, but would also likely lead to the incorporation into Google and Facebook of any successful new social media technologies, guaranteeing that the economies of scale that only these companies can bring to social media technologies would be realized. Google and Facebook would be forced to license any new social media innovations that meet with success in order to prevent their own platforms from sinking into irrelevance, so they would both pay a reward to innovators and incorporate the new innovations, giving them the greatest possible reach. You would get the benefits of size without the costs to innovation.
That would be a big improvement over breakup, which would lead to smaller and hence, for users, less valuable, social media platforms, at least until a new dominant platform could emerge, as one surely would given the value of size to any social network. But with the return of a dominant platform, the kill zone would presumably reappear and the rate of innovation would fall again. Thus in a breakup scenario, unlike in an intellectual property scenario, one cannot both have economies of scale and innovation at the same time.
Of course, any intellectual property rights regime comes with its own bureaucratic costs: someone must decide what really is a protectable social media innovation. And there is the possibility of holdup. There is never a guarantee that any two parties bargaining over a license fee will actually reach agreement, and if they don’t then the technology will not propagate. But holdup, at least, can be solved by a regime of compulsory licensing at regulated rates. (If it seems like a lot to ask a regulator to decide on a license fee, shouldn’t it seem like quite a bit more to ask a regulator to decide precisely how Facebook or Google should be broken up?)
But the really nice thing about an intellectual-property-based approach to kill zones is that it is also a diagnostic tool that can be applied in advance of taking any difficulty-to-reverse actions. We would be able to diagnose the existence of kill zones if, in response to the creation of intellectual property rights, startups were to appear and Google or Facebook were to license the startups’ technology. The fact that Google and Facebook would choose to license the technology would tell us that these big companies view the technology as a competitive threat, and that would in turn tell us that in the absence of the intellectual property rights they would have simply copied the functionality and run the startups out of business. If, on the other hand, startups do not appear or Google and Facebook do not license their technologies, we could conclude that there are no kill zones and phase out the intellectual property rights.
By contrast, under the breakup approach, there would be no way to know whether there was kill zone until after breakup, because only then could we see whether innovation had increased in response. And if there were no increase in innovation, suggesting that there had not actually been kill zones, we would need to wait for the smashed bits of the Tech Giants to knit themselves back together before we could reacquire the benefits of large networks that we currently enjoy.
Indeed, even those who remain committed to the breakup remedy ought to support the imposition of some sort of intellectual property rights regime first, purely as a diagnostic tool. Introducing intellectual property rights and seeing what happens would tell us more about whether there really is a kill zone than any economic study to which an antimonopolist might appeal for diagnostic support today, because any economic study is necessarily counterfactual: the economist can only draw on data about the industry as it is now, or about other industries, to decide whether there would be more innovation in the event that Google and Facebook were to change their behavior.
By contrast, introducing intellectual property rights has the character of an experiment designed to elicit a response (licensing of new startup innovations) that can exist only if the underlying kill zone disease is present.
Would Google and Facebook try to game the diagnosis, by avoiding licensing any technology under a temporary, diagnostic intellectual property regime in order to avoid sending the signal that there is a kill zone? I think not, because doing that would be their loss. If Google and Facebook don’t license successful new technologies, competitors will grow at Google and Facebook’s expense, and they won’t enjoy a dominant position anymore—in which case we would end up with more long-term competition in the market, which is what antimonopolists hope for anyway.
Of course, such a diagnostic experiment would hardly be the sort of large-scale, controlled undertaking one needs for scientific certainty. But it would be a start, and perhaps inch antitrust in the direction of the level of respect for the complexity of its subject that is characteristic of a science.