Thread · 7 tweets · 31 Jan 2020

It's concerning to see this analysis from @FTCPhillips that misses the structural interplay between privacy and monopoly, in this case between dominance over browsers (piggybacked off mobile OS dominance) and ads, with poor privacy in both. Thread 👇

I know first-hand that Chrome has an outstanding engineering team that cares deeply about its users, yet the browser remains engineered for tracking. Without influence from Google's ad arm, even if gentle, they would have implemented excellent privacy years ago.

Because of strong network effects in personal data, and of the non-excludable nature of data, advertising based on personal data will, absent intervention, lead to concentration. The rest of adtech tries to be a bunch of small Googles when scale alone decides returns: can't work.

Some parties however *can* make access to data excludable: browsers, mobile OSs, identity providers. If these are made to work for users rather than for an ad intermediary, they will switch to strong privacy.

In turn, this will cut off the mechanism that ensures that the digital ad market is overconcentrated: by limiting access to data that is more valuable the bigger the player.

Additionally, decreasing the concentration of digital advertising will remove the killzone around innovative advertising alternatives, including innovative features from browsers in this space.

We've made a tangled mess no doubt, but those of us who get the tech and the economics needs to do a better job telling policymakers about the central role of user agents and the effects of scale in data that make it a different commodity.