"State Regulation of Online Behavior: The Dormant Commerce Clause and Geolocation"

Jack Goldsmith's and my article with this title is now out in the Texas Law Review, as is David Post's response (many thanks for writing that, David!). Many of our readers are already familiar with it from when I serialized it here several months ago; but here's the Introduction:

Many state laws apply to internet communications. Indeed, we take many such laws for granted. If you publish an online magazine or a blog that comments on people from all fifty states, you might be subjected to the libel laws of each state. If you sell online images of famous people (or, to be au courant, non-fungible tokens), you might be subjected to each state's right-of-publicity law. Likewise as to the torts of disclosure of private facts, false light, and more. To be sure, the First Amendment uniformly protects much of this speech. But if you go beyond the First Amendment's protections, you could in principle be subject to many different state laws.

When, if ever, must courts reject such laws as unduly burdening interstate commerce in violation of the Dormant Commerce Clause? Courts in the 1990s and early 2000s often used the Clause to invalidate some internet-related state statutes—especially ones that restricted "harmful to minors" material. But more recently, and increasingly, courts have upheld state laws regulating various internet transactions.

The issue has been most notably implicated by recent state statutes that limit platforms' ability to block user posts based on the posts' viewpoint. The Florida and Texas social media platform viewpoint-neutrality statutes were indeed challenged under the Dormant Commerce Clause—and though the courts didn't reach the challenges because they struck down the statutes on other grounds, the question will doubtless recur as states increasingly seek to regulate social media platforms. The Supreme Court's decision returning abortion regulation to the states may also lead to statutes limiting abortion advertising that is targeted at states where abortion is illegal, and to Dormant Commerce Clause (as well as First Amendment) challenges to those statutes.

The Dormant Commerce Clause argument against state regulation of internet services is basically this: By imposing liability on internet speech sent to one state, a state law would potentially affect speech sent from and received in other states, and would in this respect be improperly extraterritorial. Requiring platforms or speakers to consider the laws of all fifty states can gravely burden such entities and therefore interstate commerce. And in some situations, the laws may even conflict with each other—for instance, if state A limits sending pornographic material into the state in a way that children can easily access it, but state B makes service providers quasi-common carriers that are barred from blocking such material.

Yet there is good reason to preserve state discretion here: American federalism has long embraced a territorialist-pluralist vision of different states having different laws, as the example of varying tort law rules illustrates. These differences stem in part from different states having laws that presumably match the views of their populations, which naturally differ from state to state. But even beyond that, this vision allows for experimentation, with different states testing out different rules that may then be evaluated by courts and legislatures in other states (or by Congress). Against this background, our federal system presumptively preserves traditional state power to control what happens "in" or what is sent "into" states, and to protect state residents from what the state perceives as harms.

A quarter-century ago, the internet seemed to make this vision impossible to preserve. But today, technology can enhance such territorialist pluralism. Online services can, relatively reliably, determine the state in which a user is located, and their software can then act differently depending on which state is involved. Such so-called geolocation isn't perfect; but so long as the law requires only reasonable attempts at geolocation rather than perfection, the burden on interstate commerce ought not be excessive. As the Ninth Circuit said in rejecting a Commerce Clause challenge to a California law that required CNN (among others) to provide closed captioning on programs downloaded by users in California:

[T]he DPA [Disabled Persons Act], which applies only to CNN's videos as they are accessed by California viewers, does not have the practical effect of directly regulating conduct wholly outside of California. Even though CNN.com is a single website, the record before us shows that CNN could enable a captioning option for California visitors to its site, leave the remainder unchanged, and thereby avoid the potential for extraterritorial application of the DPA….

In fact, CNN already serves different versions of its home page depending on the visitor's country and provides no explanation for why it could not do the same for California residents.

This Article explores what geolocation technology means for the Dormant Commerce Clause. We build toward an analysis of state regulations of social media platforms because they are in the news and currently being challenged in courts. But as our reasoning along the way makes plain, the analysis applies to Dormant Commerce Clause issues implicated by a much wider range of state internet regulation as well.

This Article was written and circulated for publication before the Court granted certiorari in National Pork Producers Council v. Ross, a case raising Dormant Commerce Clause issues related to our analysis; the Article is being published before the case is decided. But once the Court decides that case, we will publish an update in Texas Law Review Online (Volume 102) that will discuss how (if at all) National Pork Producers affects our analysis.