Speed for Sale

The bill died in committee. It never got a floor vote. The cable and telecom industry had more lobbyists than Markey had votes, and they were clear about the correct way to think about network architecture: the internet was not a utility, it was a product, and products should be priced according to what the market would bear, and anyone who disagreed was misunderstanding how innovation worked.

Twenty years later, the ISPs are still right. Not because the policy argument was wrong—it wasn't—but because they were correct about the power dynamics from the beginning. The fight over net neutrality has been the story of a sound structural idea running into the wall of institutional incentives every three to five years, losing in a slightly different way each time, and having the losses framed as near-victories that just needed one more administration to stick.

This is that story. It is not a story about almost winning.

i · the architecture was always the problem

The foundational issue with net neutrality as a regulatory goal was jurisdictional, and it was created deliberately. The internet was classified as an "information service" rather than a "telecommunications service" under the Telecommunications Act of 1996. That distinction mattered enormously. Telecommunications services—phone companies—were regulated as utilities under Title II of the Communications Act, meaning the FCC could issue binding rules about how they operated. Information services sat under the lighter-touch Title I, where the FCC had authority to issue guidance but not to enforce much.

This classification wasn't an oversight. Cable and telephone companies had spent most of the 1990s lobbying for information-service status specifically because it limited regulatory exposure. They understood, even then, that someone would eventually try to use the FCC to impose access rules. The classification was a hedge against that future—a legal moat dug before anyone knew exactly what it would protect.

The FCC didn't run into the wall until 2008, when Comcast was caught secretly throttling BitTorrent traffic on its network—not announcing it, not disclosing it, just slowing it down without telling subscribers. The FCC ordered Comcast to stop. Comcast sued. In 2010, the DC Circuit Court ruled that the FCC lacked authority to impose network management requirements on an information service provider. The commission had been trying to referee a game on a court it didn't own.

This created the core loop that would define the next fifteen years: the FCC would try to establish neutrality principles, ISPs would challenge in court, courts would find that the FCC's statutory authority didn't cover what it was attempting, the FCC would recalibrate, and the cycle would begin again. Each iteration consumed two to four years and cost the public tens of millions in regulatory resources while the ISPs wrote the legal fees off as a cost of doing business.

Markey introduced more bills. The Save the Internet Act. Various broadband reform proposals. None passed. The jurisdictional trap held.

ii · the high-water mark, and what followed

Tom Wheeler became FCC chairman in late 2013. He was a former cable lobbyist, which generated its own news cycle of obvious jokes. Wheeler spent most of 2014 trying to craft rules that would allow ISPs to charge content companies for "priority access"—essentially, creating fast lanes for whoever paid—while calling it net neutrality. The distinction between "net neutrality" and "managed fast lanes you have to pay for" is, to be clear, the entire question.

Four million public comments later—and after public pressure from President Obama—Wheeler reversed course. In February 2015, the FCC voted 3-2 to reclassify broadband as a Title II telecommunications service. The Open Internet Order prohibited blocking, throttling, and paid prioritization. For the first time, the internet was regulated like infrastructure.

ISPs immediately sued. The DC Circuit upheld most of the 2015 rules in 2016. It looked, briefly, like something had been resolved.

Ajit Pai became FCC chairman in January 2017. He had spent his entire tenure as a commissioner writing dissents about regulatory overreach. His first notable act as chairman was blocking a broadband privacy rule that would have required ISPs to get consent before selling subscribers' browsing data. By December 2017, the Restoring Internet Freedom Order had passed 3-2, reclassifying broadband back to Title I and eliminating the 2015 rules.

The broadband providers issued statements expressing their commitment to an open internet. They simply preferred that commitment to be unenforceable.

California passed its own net neutrality law, SB-822, in 2018. The ISPs sued. The Ninth Circuit upheld it in 2022. A small number of other states began passing similar laws. The patchwork of state rules that the ISP lobby had warned would be unwieldy and inconsistent arrived—as a direct consequence of their success in eliminating federal standards.

In 2024, the Biden FCC voted 3-2—the same margin as 2015—to restore Title II classification. The ISPs sued within weeks. They had a new tool: the Supreme Court had just decided Loper Bright Enterprises v. Raimondo, overturning the Chevron doctrine that had required courts to defer to agencies' interpretations of ambiguous statutory authority. Without Chevron, the FCC couldn't argue that its reading of its own jurisdiction deserved judicial respect. The Sixth Circuit vacated the 2024 order later that year. The FCC, for the fourth time, had its rules overturned.

iii · what the fight was actually about

Here's what the net neutrality debate contained, underneath the jurisdictional arguments: a dispute about whether the internet was a commons or a product.

Markey's 2006 bill—and every iteration of neutrality policy that followed—was premised on the internet as shared infrastructure. The underlying network had been built with substantial public investment: ARPANET, NSF grants, spectrum allocations, university networks. Private companies had invested heavily in last-mile infrastructure, but they had done so within a regulatory and investment environment shaped by decades of public resources. The implicit premise of neutrality regulation was that you don't get to discriminate on infrastructure you didn't build alone.

The ISPs operated from a different premise: capital investment creates the right to shape what the capital built. They had laid fiber and installed equipment, and the returns on that investment should be theirs to structure. If Netflix uses more bandwidth than a small blog, Netflix should pay for that difference. If an ISP wants to offer faster service to a content company willing to pay, that's a business arrangement, not a public policy problem.

Twenty years of legal and regulatory battles have resolved this dispute in the ISPs' favor, at the federal level, for now. The Supreme Court's effective withdrawal of Chevron deference has made it significantly harder for the FCC to establish any durable regulatory framework without explicit congressional authorization. And Congress, in twenty years, has passed nothing. Not Markey's 2006 bill, not its many successors, not anything that would codify what the FCC kept trying to establish through rulemaking.

The broadband lobby spent what various estimates put at over a billion dollars in lobbying and legal fees over this period. The math worked: avoid binding regulation, preserve pricing flexibility, keep the legal challenges going until the political cycle turns. It worked in 2017. It worked again in 2024. The current FCC under Brendan Carr has no interest in revisiting this.

The internet still mostly works. Pages load. Video streams. The catastrophic stratification that neutrality advocates warned about—where startups couldn't compete with incumbents who could buy priority access, where content companies faced ISP extortion—hasn't fully materialized, because the business incentives for maximum extraction haven't aligned in the worst possible way yet. The architecture is in place. The legal framework permits it. The ISPs have every tool they'd need if they decide to use it.

Markey introduced his bill in 2006 because he could see where the logic of unregulated network management led. He was correct about the destination. He was wrong, or perhaps just optimistic, about whether Congress would act before the ISPs captured enough of the regulatory and legal landscape to make acting irrelevant.

The speed was sold. We've been paying installments ever since. The balance is still outstanding.

iv · sources

• Network Neutrality Act of 2006 — Sen. Ed Markey press release, 2006-05-02
• FCC Votes to Restore Net Neutrality — Ars Technica, 2024-04-25
• Net Neutrality — Electronic Frontier Foundation
• Loper Bright Enterprises v. Raimondo — Supreme Court of the United States, 2024-06-28