Yesterday, the Ohio Attorney General filed a lawsuit asking an Ohio state court to declare Google a common carrier and/or public utility under the laws of Ohio and Ohio common law. (News release here; complaint here.) Here’s my hot take just from reading the complaint and with zero Ohio law research: It’s novel, and not obviously stupid. But it has some real obstacles to overcome.
I stress this because I expect most people will find this so mind boggling that they will be tempted to write this off. Don’t. It’s a novel application of traditional common carrier law, but that is how law evolves.
That said, I don’t think it’s a winner. But I would need to do some serious research on how Ohio common law has dealt with particular key elements of the common law, embodied in Ohio’s statute as serving the public “reasonably and indiscriminately.” Keep in mind I’m not saying that I think this is necessarily the right policy. Indeed, my colleague John Bergmayer at Public Knowledge has explained why treating digital platforms as common carriers could be a very bad idea.
A brief explanation of all this below . . . .
How Does Ohio Get to Regulate Common Carriers? I Thought That Was a Telecom Act Thing?
As I have written at length before, the history of common carrier regulation goes back 500 years in the common law. All states therefore have their own versions of this doctrine and regulate businesses as “common carriers.” As University of Indiana Professor Barbara Cherry has written about in a number of articles and FCC filings, both “common carriage” and “public utility” are categories that traditionally apply based on what you do and how you behave. A state or federal statute can override the common law, of course, just as a federal or state statute can alter the common law of negligence or anything else. But absent a conflicting statute (we’ll deal with possible conflicting Ohio and federal statutes later), a state can apply its common law or statutory definition of common carrier or public utility definition to whatever business meets that definition.
This includes interstate services, subject to the usual issues about burdening interstate commerce. Certainly Google has a Commerce Clause argument to make here. But simply because the service operates in all states does not mean that a state cannot regulate it within the confines of the state and strictly with regard to the impact on its own residents. See Head v. New Mexico Bd of Examiners, 374 U.S. 424 (1963). As I will repeat multiple times, I’m not saying who will prevail here. I’m just saying this should survive the pleading stage because it is not obviously wrong.
So Assume OH Can Apply Its State Law, How Does Google Act As a Common Carrier?
Traditionally, common carriers have the following qualities: they do business with the general public and treat all members of the public “indiscriminately” or “indifferently.” That means that they don’t engage in individualized contract negotiations when providing a service or limit themselves to some group like commercial customers. So for example, a contractor who comes to fix up your house is not a common carrier. Yes, they advertise to the general public, but they usually insist on getting details about a job to determine if they can do it, and then negotiate a rate for the specific job.
Many states also require that the business somehow be “affected with the public interest.” That can mean a bunch of things. Usually it refers to the importance of the service. For example, among the oldest common carriers are “carters” and “teamsters,” because hauling goods from place to place is essential to commerce. This is why things like delivery services, taxi cabs, and other competitive businesses are regulated as common carriers. It’s not a question of competition (although some cases, such as Munn v. Illinois, do reference the “quasi-monopoly” quality of many common carriers subject to franchise). A service can be highly competitive and still meet the definition of common carrier.
So Is Google A Common Carrier?
Excellent question that requires specific research under Ohio law. My gut is that because Google makes carefully differentiated choices based on the individual searching, the nature of the query, and whatever other factors go into its algorithm, that Google is making individualized determinations that are the antithesis of common carriage. OTOH, OH can argue that Google is putatively applying the same algorithm to every search. So while the results may be unique to each individual, Google is treating all members of the public the same by applying the same algorithm.
If this sounds familiar, it is similar to the question that was originally raised in Masterpiece Bakeshop v. Civil Rights Commission of Colorado, aka the “gay wedding cake case.” That case dealt with whether the bakery was a place of public accommodation (which uses similar test to common carrier for purposes of this comparison) and therefore subject to Colorado’s civil rights law — which prohibits discrimination based on sexual orientation. The baker argued that when he makes wedding cakes, he is creating individual works of art/food and therefore was not dealing with the public indiscriminately. Colorado argued that minor variations when making a wedding cake or other decorative cake does not change the fact that the bakery generally does business with the public and that the process of making cakes is sufficiently standardized as to count as dealing with the public indiscriminately. Ultimately, the Supreme Court reversed on other grounds, so we don’t have a good answer here. But it gives you a sense of why this question is generally more complicated than a simple yes/no.
Assume Google Meets the Definition Under Common Law, What Other Defenses Can Google Raise?
Other than the aforementioned Commerce Clause defense, Google has three primary defense lanes that I can see with virtually no research and no factual record: state law defenses (a state statute that says Google can’t be a common carrier), federal preemption (which generally focuses on Section 230 and the Communications Act generally), and the First Amendment. I will take each in turn.
Possible State Law Defenses.
Let me start by noting that the complaint does not provide any citations for its assertions under OH common law (although it does provide citations for the jurisdictional issues to keep the case in state court). That makes it hard to assess the stength of the arguments made. I’ll also remind folks this is complicated and I’ve done zero research into OH law, so this is something of a hot take and Google with its cadres of lawyers will undoubtedly come up with lots of state law defenses I don’t talk about here.
But thanks to tech law Twitter, and this tweet by U Chicago Law Professor and fellow Princeton Alumn (go Tigers!) Genevieve Lakier, I can point to one potential OH law problem. Section 4095.02 of the OH Revised Code, “Public Utility Defined,” excludes certain services from the definition of “public utility.” Of relevance here, 4095.02(A)(5)(c) excludes any provider of any “Information service as defined in the “Telecommunications Act of 1996,” 110 Stat. 59, 47 U.S.C. 153(20).” Google Search pretty clearly fits the definition of an “information service,” so this would seem to exclude Google from being a public utility in Ohio. Whether that automatically precludes it being a common carrier under Ohio law I have no idea. But this naturally takes us to the next line of defense, federal preemption.
What About Federal Preemption? Isn’t This Being Litigated in the Net Neutrality Context?
Yup. In California, ISPs have gone so far as to argue that the Communications Act of 1934 preempts any interstate information service because it preempts every form of interstate communications. I’ve written about why I don’t think the Communications Act generally bans regulation of any “interstate information service.” I’ll add that Section 414 of the Communications Act provides a general savings clause for any state or common law remedies that don’t expressly conflict with the Communications Act. Seeing as how this is a common law remedy, unless there is a specific conflict then OH law is not preempted (by fed law, OH may have preempted itself as discussed above).
Which brings to Section 230, because no conversation about regulating digital platforms is complete without bringing up Section 230. The complaint cites Justice Thomas’ concurrence in Biden v. Knight First Amendment Institute of Columbia University, which itself relies heavily on this article by Adam Candeub, that essentially argues that one can regulate access to speech of common carriers and that digital platforms could be classified as common carriers. But that doesn’t tell us whether or not a federal law preempts state law. All that says is that it is theoretically possible to justify application of common carrier regulation to digital platforms.
So we turn, as always, to the text of Sec. 230. Section 230(c)(1) prevents treating a digital platform as a publisher or distributor “for any information provided by a third party.” Applying the plain language of the statute, treating a digital platform as a common carrier does not violate this provision. Indeed, it does the opposite. The common law (and state and federal law) generally exempt common carriers from liability for any third party content published or distributed through their system because they have no choice about carrying it. But the analysis does not end here, as some courts have broadly interpreted Section 230(c)(1) to mean that any time a regulation or common law obligation tries to make an interactive service do anything specific with regard to third party information, this is attempting to treat the interactive service as a publisher. I have no idea where the Sixth Circuit comes out on this. Compare Herrick v. Grindr (attempt to apply product liability law rejected as effort to hold service liable for 3rd party content in violation of Section 230(c)(1)) with Fair Housing Council of San Fernando Valley v. Roomates.com (Section 230 does not apply when interactive service actively solicits information used to violate civil rights law). Even where Section 230 provides some protection, it does not necessarily shield the interactive service provider from state law entirely. See Oberdof v. Amazon (Amazon subject to state product liability law as “merchant,” but not obligated to post a warning about the product because requiring Amazon to post specific content would interfere with its editorial discretion).
Additionally, Section 230(c)(2)(A) prohibits any liability for removing or restricting access to third party content that the provider considers “objectionable” for any reason, provided the action is taken in “good faith.” While courts frequently conflate the Section 230(c)(1) and (c)(2) analysis, this provision has also been given very broad scope in allowing providers of interactive services to block or otherwise limit access to third party content. See Zango, Inc. v. Kapersky Lab, Inc. (Cybersecurity software protected by Section 230(c)(2) for blocking adware as “otherwise objectionable”).
Given all this, it is easy to see how even if it is theoretically possible to regulate a search engine such as Google as a common carrier, application of any such state law claim is blocked by Section 230. I’m inclined to lean that way based on my current state of knowledge — which includes a complete ignorance of the relevant 6th Circuit precedent and a good technical explanation by either party as to why application of Section 230 (either relevant provision) does or doesn’t apply here. As I keep saying, I’m not trying to argue which side is likely to win or even which side has the better argument. I’m just saying that Ohio’s complaint is plausible enough to not be stupid, despite being a novel application of the law.
What About the First Amendment?
Lord knows this gets brought up whenever we talk about regulating commercial practices that touch on speech. But the First Amendment is not the total regulation blocker that some folks (including some judges, unfortunately) seem to think it is.
To start, as Thomas points out in his Biden v. Knight 1st Amendment concurrence, “regulations that might affect speech are valid if they would have been permissible at the time of the founding.” This comes from one of the Court’s First Amendment absolutists. Basically, when common carrier is applicable, First Amendment for the common carrier basically drops out. The First Amendment right applies to the speaker, not the carrier. See Sable Communications v. FCC, (attempt to regulate “dial-a-porn” violates First Amendment rights of users of telephone service). So if Google is a common carrier, the non-discrimination obligation comes with the package without violating the First Amendment.
But even if we apply the First Amendment, we are likely to apply intermediate scrutiny rather than strict scrutiny. This is, after all, economic speech under Central Hudson. A non-discrimination obligation is a classic content-neutral obligation. This isn’t about asking Google to police speech based on content, but to avoid discriminating based on economic factors (here, self-preferencing for anti-competitive advantage). Additionally, such a non-discrimination requirement is no more restrictive than is necessary to achieve the government purpose (here, promoting competition in the state of OH). So if OH can get past the other hurdles, I think they prevail on a First Amendment defense as well. But again, I’m just speculating based on the complaint and general principles of law rather than on any case-specific research.
It is easy to dismiss the OH AG complaint against Google as yet another headline-grabbing Republican official pandering to the loathing of their base for “Silicon Valley” companies. It is also easy to reject that idea that a 500 year old common law doctrine designed for folks pulling carts of vegetables or wool to London applies to a modern search engine. But law evolves by taking traditional principles and apply them to new circumstances in novel ways. Something that one its surface looks ridiculous can make a good deal of sense if you strip away the surface differences and get to the core principles at issue. In this case, the law of common carriage applies to very real modern questions about whether a business essential to commerce has a legal obligation to conduct itself fairly, without privileging itself or otherwise abusing its position.
This doesn’t always work, or always make a particularly good fit. But in the absence of any federal action, states are going to take the steps they deem necessary to protect their residents or otherwise achieve what they think are the right public interest goals. I can’t say that OH is going to win, or even judge if they have a good case. But this isn’t like the effort to get the FCC to regulate social media companies. Whatever the motivation, there is nothing obviously crazy about the OH complaint. We will have to see what Google has to say in response, and what the OH state court ultimately says.
Stay tuned . . . .