Why does American broadband still have block-to-block lottery conditions

I'd predict we see more of the "municipal fiber as a workaround" pattern over the next decade, but it'll stay patchwork and won't actually solve the structural problem. Cities with political will and some cash (usually mid-sized college towns or wealthy suburbs) will build out their own networks. They'll work fine. Then you'll hit the boundary and drop back to Comcast, and everyone will point to it as proof that municipal broadband is inefficient, rather than proof that the fragmentation itself is the inefficiency. The federal level won't touch the monopoly question because the real political cost isn't economic—it's that you'd have to pick a loser. Comcast doesn't need to cooperate with a breakup. They just need enough of Congress to decide the status quo is fine, and that's always an easier ask than the alternative. We're more likely to see another round of "broadband subsidies" that flow money to the existing providers to upgrade their own territory, which is basically rent extraction with a public funding mechanism attached. What might actually move it: if some state gets frustrated enough and just nationalizes the last mile aggressively, and it doesn't immediately cause the apocalypse those companies claim it will. But that requires a state to absorb the political heat for several years before anyone can say whether it worked. Hard to see who takes that bet first.

I've watched this play out from the infrastructure side, and you're nailing something important but I'd push back slightly on the "nobody wants the fight" part. The fight exists—it's just asymmetric as hell. We had a similar situation at my last place. Building needed actual redundancy, not just speed. The incumbent's response to "we'd like another provider's fiber in this conduit" was basically a shrug and then six months of permitting delays that only they could waive. They didn't say no. They just made it technically someone else's problem. By the time you get through the pole attachment disputes and the "we need to inspect the conduit" theater, you've spent enough money and time that you pick up the phone and negotiate with the incumbent instead. The system works exactly as designed for them. What I realized recently though: the actual barrier isn't even the regulation anymore. It's that once you've got fragmented ownership of the physical plant, the cost of fixing it becomes so enormous that it actually *does* become a technology problem in practice. You'd need to either overbuild everything (which kills unit economics for anyone but a VC-backed startup in a rich neighborhood) or force wholesale access at prices the incumbents will fight to the death. The regulation created the mess, sure, but now the mess is self-reinforcing in a way that's almost independent of what you change about the law. Europe paid that cost up front. We're paying it now, slower, in a worse way.

You're describing the right structural problem, but I'd push back slightly on the "accidentally locked in" framing. The 1996 Act didn't accidentally do this—it was a deliberate choice to let incumbents keep their territorial moats in exchange for some deregulation on other services. Regulators knew what they were doing. It just turned out to be a bad trade. The clearer distinction: you're separating the *technology problem* (fiber is expensive to deploy) from the *incentive problem* (no one with control of the pipes has reason to deploy it). Europe didn't solve the technology problem—fiber still costs money to bury. They solved the incentive problem by making the infrastructure a public good instead of a private protected asset. That's not regulation preventing competition, that's regulation *enabling* it by treating the pipes like pipes rather than like proprietary product. The reason this matters for your "nobody wants the fight" point: every proposal to actually fix this (open access mandates, forced unbundling, public fiber buildout) requires someone to take away what ISPs already own or control. And they've spent 25 years turning that control into stable revenue. It's not a technology problem you need to convince Congress about. It's a taking problem, and those are much harder to sell.

The distinction you're drawing here is important but slightly off. It's not just that they didn't have to share infrastructure—it's that the 1996 Act created a specific asymmetry: it allowed cable and telcos to maintain exclusive control of the physical plant (the ducts, poles, conduits) while theoretically opening up service competition on top of it. Except that exclusivity over the physical plant is what actually matters. You can't build a competing network without access to those conduits, and the incumbents had zero obligation to lease space on reasonable terms (or at all). The result is what you're describing, but it's worth being precise about the mechanism: the law didn't just fragment the market, it enshrined a particular kind of fragmentation where infrastructure ownership equals service monopoly. Europe's approach was messier in practice—some countries did this badly too—but the ones that worked treated last-mile conduits as a common resource separate from the service layer. That's a much harder thing to undo than just "breaking a monopoly." You have to create a whole new property regime first, and existing owners will fight that harder than they'd fight price regulation. Your building's situation is the purest version of this problem. Two blocks away there's perfectly good fiber, but the ducts connecting your block to anything competitive don't exist and won't be built because the incumbent already owns the path of least resistance. That's not a technology problem or even a regulation problem in the sense people usually mean—it's a property rights problem baked into the physical world.