The big five tech firms, Facebook, Amazon, Apple, Netflix, and Google, known more commonly as FAANG, have a market capitalization of $5.6trn, almost a fifth of the S&P 500 Index. Naturally, regulation of such a behemoth of the American economy is a thorny issue because these big tech firms create jobs, spearhead American innovation, and reinforce America’s position as a world leader. Before the rise of big tech, monopolies like Bell Labs were regulated to prevent them from abusing monopoly power through bundling, predatory pricing, and vertical restraints. This all amounted to, in one way or another, a great asymmetry between consumer and monopoly surplus. Today, these tell-tale fiduciary signs of abuse of monopoly power are sometimes transparent. It is not easy to see how predatory pricing could befall customers if services like Facebook Messenger and Instagram and Google are free. Viewed from the lens of a data economy however, we see that the big five are not simply large internet providers anymore; they are a colossus with a long reach in every corner of our lives, including politics, retail, national security, and privacy. Regulation in the case of big tech is needed to protect the data privacy of users. Without regulation, big tech can create a “bubble” around a user, not allowing them to even know about competitors, giving the existing tech monopolies free rein over a user’s life choices, and collectively, even some social and political outcomes.
To analyze regulation of “data” companies like Facebook and Google (that get the majority of their revenue from targeted ads using personal data), we can’t use traditional methods of looking at the inverse demand function, because the services are free to use. We could try to use a proxy and measure a “willingness-to-pay” in terms of amount of personal data users are willing to give up in order to continue using the service, but it would be difficult to measure deadweight loss in this scenario. Instead, what we could do is measure the average marginal revenue brought in by one user’s personal data in terms of targeted ad revenue.
If our goal is to restrain the impact big tech has in the non-tech realm, then to regulate their targeted ads business would be to chip away at their reach in other industries. As Paul Romer suggests in his 2019 New York Times article, a sales tax on the revenue earned by a company for displaying targeted ads would force tech companies to innovate or be subject to competition, who don’t have the data for targeted ads yet. However, it might also be interesting to bring the tax down one more level to taxing companies for additional data over some specified amount. As in Averch-Johnson’s model of the firm under regulatory constraint, the regulator knows the firm’s capital, and chooses some fair rate of interest, which then determines the firm’s profit. In my proposed tax, user’s private data, which would be used to hone machine learning algorithms and produce targeted ads, would be considered capital. In this case, however, the regulator would need to determine a “fair threshold of data” which would allow the firm to make a profit, and any excess data would be taxed. This would allow competition because smaller companies would not have enough data to reach the threshold. It would also force companies to innovate: some firms might not collect additional data of users and accept the concomitant reduction in targeted revenue ads; some firms might seek better ways to draw the same conclusions but from less data or noisier data.
In the past year, FAANG’s combined value increased by over $2trn – a bull run that has stemmed from their ability to construct an individual world for every person. As these individual worlds get more and more personalized, the firms’ ability to manipulate the individual, and thus the masses, also becomes more apparent. A legislation to prevent that would check that power, and begin to restore the ability for each person to create their own, recommendation-engine-free, reality.
Peter's Life 