Is the marketplace of ideas broken? My colleagues David Pozen (Columbia), Yochai Benkler (Harvard), and Amy Kapczynsky (Yale) all seem to think so. I already dispatched Pozen and Benkler’s proposed cures — Pozen wants more big tech censorship (see here), while Benkler (here) wants more “social democracy” (like Trudeau’s Canada, I presume?) — and I also presented Professor Kapczynsky’s more nuanced package of remedial proposals in my previous post. In her ideal world, or if she were “Queen of the Internet” for a day, she would (1) require Big Tech firms to publicly disclose their algorithms, or what she calls “data publicity”, (2) she would break up Big Tech firms into smaller companies (“regulate their size”), and (3) she would even prohibit Big Tech firm from making money from their “data gathering” activities.
[Note: Kapczynsky also favors the creation of “public data trusts” and wants to “figure out new ways to support public media and intermediaries” — i.e. she wants the government to run its own social media platform. (You can read her entire essay for yourself here.) But this last proposal is so preposterous on its face that I won’t even bother refuting it here, except to ask a simple rhetorical question: Why does Kapczynsky trust the government with user data more than she trusts Big Tech?]
What about Professor Kapczynsky’s other Internet proposals, “data publicity” and anti-trust regulation? Read on:
- Data publicity. The most glaring problem with Kapczynsky’s call for “data publicity” is that, with all due respect, she has not really thought through the likely real-world adverse consequences of this proposal. “Data publicity” may sound great on paper, but what effect would this policy (making data algorithms public information) have in practice? Alas, I suspect that data publicity would most likely give incumbent Internet platforms like Facebook, Google, and Amazon (the very firms that Kapczynsky deplores) a major advantage over future competitors by making it more difficult for new startups to enter the Internet platform market! After all, if new ventures were required to make their data algorithms public, then those algorithms could then be used by incumbent firms to fend off new entrants.
- Size and data regulation. Kapczynsky further claims that the size and business practices of Big Tech firms like Google and Facebook must be regulated because those firms are too big, control massive amounts of sensitive user data, and have too much influence over the Internet advertising market. Sigh. Aside from committing the Nirvana Fallacy (see here, for example), Kapczynsky’s call for anti-trust regulation has two fatal flaws. One is that regulations are usually counter-productive. They increase the costs of doing business and thus end up stifling competition by making it more difficult for new entrants to enter the market. The other problem is that Google and Facebook are not really monopolies, for as I explained last fall in this post, if this were true, we would expect to see fewer online ads overall, which would be a good thing for most of us end users! (Why would we expect to see less ads if Google and Facebook were real monopolies? Because demand curves slope downwards: when monopolists sell ad space at higher than competitive prices, ad agencies buy fewer ads.)
To conclude, Kapczynsky calls out at least three Big Tech firms by name in her essay (see here): Facebook, Google, and Amazon. But how can three business firms, no matter how big or powerful, constitute a “monopoly”? This alleged monopoly is either a three-headed monster or non-existent, or as my colleague and friend Mark Lemley has observed, “it’s a bit awkward to point to multiple ‘monopolists’ in a single market”! (See Lemley, “The Contradictions of Platform Regulation”, Journal of Free Speech Law (2021), pp. 313-314, emphasis and exclamation point added). Perhaps the marketplace of ideas is not broken at all. Perhaps the supposed problem of misinformation is being overblown. I will consider those possibilities and conclude this series in my next post …
Pingback: Global reply to Pozen, Benkler, and Kapczynsky: the optimal level of misinformation is not zero | prior probability