Corbin K. Barthold is Internet Policy Counsel with TechFreedom. He previously served as Senior Litigation Counsel at Washington Legal Foundation. Mr. Barthold was counsel for TechFreedom on a joint TechFreedom-WLF amicus brief filed in New York State Telecom Association v. James. This post originally appeared on Mr. Barthold’s Substack, and is reprinted here with permission.


If you follow broadband regulation at all, you know that, when he led the Federal Communications Commission, Ajit Pai axe-murdered the internet. In 2018 he pushed through the Restoring Internet Freedom Order, a fascist diktat that brought about nothing less than digital serfdom (to borrow one senator’s phrase). After that, people started getting the internet one word at a time. Good thing John Oliver ridiculed Pai on late-night television. Good thing irate protestors denounced Pai in the streets. Yes, a man went to prison for threatening to kill Pai and his family—but Pai was breaking the internet! Desperate times, desperate measures . . .

Does this all sound a little off?

Let’s start over.

In 2015 the FCC issued the Open Internet Order, which shifted broadband’s regulatory status, under the Communications Act of 1934, from Title I to Title II. Title I imposes light regulation on information services, while Title II imposes heavy regulation fit for common carriers. The internet was regulated as a Title II service from 2015, when the FCC issued the Open Internet Order, until 2018, when it issued the Restoring Internet Freedom Order. The internet was not broken before 2015, and it did not break in or after 2018. “Digital serfdom” there has never been. Pai was on the receiving end of a fit of progressive mass hysteria.

A few weeks ago, the FCC adopted a new Open Internet Order that returns broadband to Title II. Precisely because the Title I regime in place since 2018 has not broken the internet, the effort to move back to Title II has attracted little attention. Even the FCC itself seems to have forgotten what all the fuss was about. Its new order largely ignores the old cause of protecting “net neutrality”—something that’s never really been threatened—and trots out instead concerns about national security, consumer privacy, and public safety. As Commissioner Brendan Carr notes, in dissent, these new justifications can’t withstand even casual scrutiny. Not for nothing is the FCC only just now “discovering” them. The agency is reaching for any rationalization it can find. For “Title II,” as Carr writes, “is now a matter of civic religion for activists on the left.”

While the well-worn Title I-Title II debate played out once again at the FCC, something momentous occurred in New York. People barely watch what the federal government is doing about broadband anymore, never mind what’s happening in the states. But while almost no one was looking, New York enacted the Affordable Broadband Act. The ABA declares broadband an “essential service,” and orders broadband providers to supply it to low-income consumers for $15 a month. As the FCC crawled its way to revoking the Restoring Internet Freedom Order and restoring the (revised) Open Internet Order, New York went ahead and imposed ex ante rate regulation—a measure that goes beyond anything the FCC has ever pressed for.

Hey, when have price controls served anyone wrong? So let’s just nod our approval at the virtue and compassion of New York’s state legislators. Who am I to question their wisdom. But what I can’t understand is this: why isn’t the ABA preempted by federal law? Last month, the U.S. Court of Appeals for the Second Circuit ruled that the Restoring Internet Freedom Order does not foreclose enforcement of the ABA. A 2-1 panel majority confidently asserted the absence of preemption. On my reading of the decision in New York State Telecom. Assoc. v. James (2d Cir.), though, the topic of broadband regulation has simply triggered another bout of progressive gaslighting. (The Restoring Internet Freedom Order remains in effect. Sure, it might go away in a couple months, when the new Open Internet Order is set to kick in. Then again, it might not. I’ll return to this point briefly at the end.)

The ABA runs into both “field” and “conflict” preemption. Field preemption arises because the Communications Act of 1934 “occupies the field” of rate regulation of interstate communications services. After all, the 1934 Act covers “all interstate and foreign communication by wire or radio.” Conflict preemption arises because the ABA thwarts the federal policies behind both the Restoring Internet Freedom Order and the 1934 Act, as amended by the Telecommunications Act of 1996. So concluded the district court, in an order blocking the ABA from taking effect.

In considering conflict preemption, in particular, the district judge found New York’s position faintly ridiculous. The Pai FCC wanted to protect broadband from common-carrier regulation, including rate regulation. When it (mostly) affirmed the Restoring Internet Freedom Order, the D.C. Circuit agreed that (re-)classifying broadband as a Title I service was a rational way to pursue that goal. In New York’s view, the district judge observed, both the FCC and the D.C. Circuit “profoundly misunderstood” what the FCC had done. “Instead of protecting broadband internet providers from common carrier treatment and its attendant threat of rate regulation,” the FCC “actually expose[d] them to fifty states[’] worth of such regulations.” And instead of giving a thumbs-up to a rational agency order, the D.C. Circuit actually let slide an arbitrary and capricious one. Under New York’s theory, the D.C. Circuit missed that the FCC was claiming to move “away from public-utility style regulation” while in fact “causing more public-utility style regulation.”

The Second Circuit panel majority didn’t just reverse the district court; it served up a little lecture. Telecom firms “vociferously lobbied the FCC to classify broadband internet as a Title I service,” the majority piously announced, yet they “now ask us to save them from the foreseeable legal consequences of their own strategic decisions.” Foreseeable! The only thing foreseeable about the panel majority’s opinion is that, when in court, the telecom firms will on occasion get a surprise kick in the shins. How, exactly, those firms are supposed to plan around this fact is unclear.

Where did the Second Circuit go wrong? Let’s stick with conflict preemption. The Telecommunications Act of 1996 is a deregulatory statute. It describes itself as “an act to . . . reduce regulation,” the better to “encourage the rapid deployment of new telecommunications technologies.” In explaining conflict preemption, the Supreme Court has said that a state law may not stand “as an obstacle” to the “execution of the full purposes and objectives of Congress.” Here is one such purpose: under the 1996 Act, “it is the policy of the United States” to “preserve” the “free market . . . for the Internet . . . unfettered by Federal or State regulation.” The FCC’s Restoring Internet Freedom Order honored this deregulatory statute. New York’s ABA, with its price controls, defies it—it conflicts with it, and is therefore preempted.

The Second Circuit panel majority mistook the 1996 Act for a pro-regulatory statute. The 1996 Act grants the FCC the power to forbear from enforcing discrete Title II obligations on a Title II service. When the FCC does so, the majority observed, “the states are prohibited from imposing that same obligation on the [Title II] service.” And because there is this express preemptive authority in Title II, the majority reasoned, there must be no implied preemptive authority in Title I. The majority assumed that, to preempt state laws, the FCC must ratchet a service up, from light-touch Title I regulation to heavy-handed Title II regulation, and then trim Title II rules one-by-one using forbearance. But as the district court pointed out, this reading of the statute is perverse. When the FCC classifies a service as a Title I service, why should that drastically expand the universe of regulations, via the states, to which the service can be subjected? The whole point of bestowing Title I status is to head off heavy-handed regulation, be it federal or state.

The majority wondered how Title I could “confer implied preemptive authority when it does not confer express preemptive authority.” The simple answer is that express preemption and implied preemption aren’t the same thing. Title I may not contain an express preemption provision, but that does not mean the states are free to undermine Congress’s objectives, enacting laws that nullify Title I by negating its effectiveness.

The majority relied on Mozilla Corp. v. FCC (D.C. Cir. 2019), the decision that upheld (most of) the Restoring Internet Freedom Order. That is indeed where the trouble started, as we’ll see. But first I should mention that Mozilla actually points up the distinction between express and implied preemption. Mozilla found that no “express statutory authority” permitted the FCC to “categorically abolish all fifty States’ . . . authority to regulate intrastate communication.” Meanwhile, though, Mozilla did “not consider whether” the Restoring Internet Freedom Order had “preemptive effect under principles of conflict preemption.” If a party can “explain how a state practice actually undermines” that order, Mozilla concluded, “then it can invoke conflict preemption.”

By conflating express and implied preemption, the panel majority reneged on Mozilla’s pledge that states may not “actually undermine[]” light-touch Title I regulation. In fact, the majority read Mozilla as saying that Title I has no implied preemptive effect. Quite simply, the majority erased the distinction between the express preemption that was at issue in Mozilla, and the implied (conflict) preemption that was left unaddressed in Mozilla but that’s at issue here.

To see how far the panel majority strayed from fundamental principles of preemption law, consider Geier v. American Honda Motor Co. (2000). The Department of Transportation issued a vehicle safety rule, under a federal traffic safety statute, that gave auto manufacturers a range of ways to place passive restraints in their vehicles. The plaintiffs sued an auto firm for failing to use airbags, specifically. The Supreme Court held that the suit was not preempted by any express provision in the underlying statute, but that it was preempted impliedly for conflicting with the objectives of the statute and of the agency’s rule.

That is exactly the situation here. The Supreme Court “distinguishes between ‘express’ and ‘implied’ pre-emptive intent”—yet the panel majority blended them together. Indeed, the majority failed to place “weight upon [the FCC’s] interpretation of [the 1996 Act’s] objectives and its conclusion, as set forth in [the Restoring Internet Freedom Order], that a [state statute] such as this one would stand as an obstacle to the accomplishment and execution of those objectives.” That quote, so neatly adapted from Geier, illustrates the parallel (missed by the panel majority) between Geier and a case like this one. By contrast, the Eight Circuit understood the similarity when, citing Geier, it stated flatly that “any state regulation of [a Title I] information service conflicts with the federal policy of nonregulation.”

In creating Title I, what did Congress achieve? We know this much: it wanted to reduce regulation, so that information services could flourish. Well, then, was Congress fine with the idea of states imposing a patchwork of price controls—and other common-carrier rules, such as market entry and exit requirements—on information services? To fulfill Congress’s deregulatory aims, must the FCC ignore Title I (rendering it superfluous?), place services under Title II, and then fundamentally rewrite Title II through sweeping forbearance? (An approach that would create problems of its own since, under the Constitution, only Congress should be in the statute-writing business.) Or perhaps—just spit-balling here—Congress set up a scheme of light-touch regulation, in Title I, and it meant what it said? Occam’s Razor should take us a long way here.

By moving broadband “outside of . . . Title II,” the panel majority claimed, the FCC “surrendered the statutory authority” to preempt state common-carrier regulations. The FCC’s authority to preempt state law must, this thinking runs, be coextensive with what the FCC itself has the authority to do. The majority picked up this novel “asymmetry” theory—under which federal regulatory schemes can pack implied preemptive power, but federal deregulatory schemes cannot—from overbroad language in Mozilla. The asymmetry theory says, in essence, that when it comes to deregulation, preemption can only be express.

This is not sound preemption law; it’s just a prejudice against deregulation. “A federal decision to forgo regulation in a given area,” the Mozilla dissent said, quoting the Supreme Court, “may imply an authoritative federal determination that the area is best left unregulated.” The James dissent, in the Second Circuit, agreed. The FCC’s finding that broadband should not be subject to Title II regulation, it wrote, quoting another Supreme Court decision, “takes on the character of a ruling that no such regulation is appropriate or approved.” To demand more “statutory authority,” beyond the FCC’s power to place a service under Title I to begin with, is to demand that, when deregulating, and only when deregulating, Congress add: “And we really mean it.” Which amounts to an erroneous claim that federal deregulatory efforts require express preemption. The Second Circuit panel majority negated the Restoring Internet Freedom Order, and upheld the ABA, by stacking the deck against Congress’s deregulatory objectives.

Granted, there’s that new Open Internet Order floating around. If and when that order takes effect, the Second Circuit panel majority’s failure to grasp the preemptive power of Title I will cease to matter, at least as to broadband (for now). But the majority’s error will remain pernicious all the same. Its opinion isn’t just a loss for the telecom firms; it could also do great harm to email, text messaging, business communications platforms, cloud-computing services, and video-conferencing apps.

After all, email and text messaging are Title I services. And business communications platforms, cloud-computing services, and video-conferencing apps look like Title I services as well. When you think of Title I, you should think of the light-touch regulatory framework that all of these services (and their successors, such as the metaverse) need to thrive. Nothing in logic enables a state to say that broadband’s Title I status opens the way to state rate regulation, but that email’s or text messaging’s Title I status does not. Under the panel majority’s reasoning, therefore, a state might now be able to apply rate regulation to every service that is, or that could plausibly be, a Title I information service.

Recall the panel majority’s belief that the forbearance power, under Title II, is the only preemptive power in the FCC’s Title I-Title II arsenal. When it declared email a Title I service, the FCC was (according to the panel majority) giving the states a green light to rate-regulate email (perhaps by requiring providers to pay users for data). Likewise, were the FCC to declare that business communications platforms such as Slack are a Title I service, that would (according to the panel majority) give the states a green light to impose price controls on those products.

Once one remembers that Title I is about much more than broadband, it becomes absurd to assume that states get a regulatory red light when something is a Title II service, but a regulatory green light when it’s a Title I service. Such an assumption would have to stand on a further assumption that every service must be subject to heavy-handed regulation by someone. But the 1996 Act (and common sense) tells us that that can’t be right.

In the panel majority’s view, the FCC must retain for itself the power to impose price caps, by placing a service under Title II, if it is to stop states from imposing price caps. But when we pan out, and think about more than broadband, that claim looks like pure folly. Imagine that a state says it will start imposing market entry and exit rules, or pay-for-data requirements, on email. Does that mean that the FCC must move email to Title II, and then forbear from treating email like a common carrier? Why would Congress, which wants the internet to remain “unfettered by state . . . regulation,” require such a Rube-Goldberg-esque process to head off state regulation? Such a protocol would make a mockery of Congress’s conclusion that Title I is the proper home of services in need of light-touch regulation.

So the Second Circuit’s opinion in James creates a mess, even if broadband lives out its days as a Title II service. Which it probably won’t. The new Open Internet Order is likely to be either struck down in court (as I’ve explained elsewhere) or rescinded the next time the GOP controls the FCC. If and when broadband’s proper Title I status is restored or affirmed, the panel majority’s contrived regulatory “Gotcha!” will pop back to life. The great pranking of Ajit Pai will resume. And on it will go, unless and until the Supreme Court enforces Title I against the lower courts and the states.