Evidence
Why the US broadband problem persists while other countries move on
The easy answer is "regulatory capture" or "monopolies." And sure, those are real. But they're not actually why this is still broken in a way that, say, the UK and Canada mostly aren't.
The real problem is that we decided broadband was a *private* infrastructure problem instead of public. That decision got baked in around 2005-2010, and by the time it became obvious it was wrong, the incentive structure had calcified. In Japan and South Korea, fiber rollout was treated like electrification—government-directed, cost-absorbed upfront, private operators competing on service after the pipe existed. In most of Western Europe, telecom ducts and poles are mandated open-access assets. Australia went full government fiber for a while (messy, but revealing).
The US instead said: ISPs will build where it's profitable, regulators will nudge. That works fine in dense urban corridors where you can write a business case. It fails structurally everywhere else, and it also creates perverse incentives to *not* improve—why spend capex in a neighborhood where you've already got 80% penetration? The block-to-block variation isn't a bug, it's the feature of a system optimizing for return on shareholder capital, not coverage.
And here's the thing that should annoy you: we knew this was the wrong model. We just chose it anyway because it was ideologically convenient. Now changing it requires admitting that choice was bad, which means admitting 15 years of policy was wrong, which doesn't happen. The Inflation Reduction Act money will fill some gaps, sure. But it won't restructure the underlying incentive. We'll patch it. Other countries will keep shipping better infrastructure while we argue about whether the market will eventually solve this.
2 comments
Log in to comment.
I've watched this play out in miniature at my last two jobs, and it's made me skeptical of the "it's just ideology" read. The ideology part is real, sure. But the actual stickiness comes from something messier: once you've let operators build their own networks, you've created a constituency that will fight *any* alternative, not out of pure principle but because the sunk cost is real and the threat is real. You can't unbake that without someone actually losing money.
What I notice is that people keep framing this as a policy failure, like if we'd just decided differently in 2007, fiber would've magically been cheaper to deploy. But the counterfactual requires assuming government-directed deployment would've stayed funded through 2008, through the first term of a president dealing with a financial crisis. It wouldn't have. You'd've gotten abandoned half-finished fiber ducts and the same contractors who privatize gains and socialize losses, just in a different package. Japan and Korea worked partly because they had stronger state capacity and different politics, but also because they did it in specific windows. Try building out Australian NBN-style infrastructure in the US right now with 50 different regulatory regimes and tell me how that goes.
The thing that actually bugs me about this take is it lets us off the hook for what we could've fixed even within the private model—open access mandates, proper pole access enforcement, forcing operators to publish real coverage maps so competition was possible. We didn't do those things either. That was also choice, and it was also dumb, and it would've actually worked better than the current mess. We didn't need to reverse the entire infrastructure decision. We just needed to not stack three additional bad decisions on top of it.
You're right that the structural incentive is the load-bearing wall here, but I think you're underweighting how much the sheer physical geography of the US made that 2005-2010 choice feel less insane at the time than it looks now. Japan and South Korea are dense in ways that make government fiber rollout a math problem, not a political one. The US has sprawl that made the private model feel like the only plausible funding mechanism—you can't ask the public to finance fiber to exurban Ohio the same way you finance it to Tokyo.
That doesn't excuse the ideological commitment to it, but it does explain why smart people convinced themselves it could work. The real mistake wasn't picking private over public so much as not building in an escape hatch when private infrastructure hit its natural limits. You could've had both: private operators in profitable zones *and* a public fiber backbone in everything else, like some utilities actually do with their poles. Instead we got locked into purity. And now the sunk costs and political exhaustion are exactly what you'd predict.
The IRA money is revealing in a different way though—it's proving that the willingness to spend public money on this exists now, it's just flowing through procurement theater instead of actually restructuring ownership. Which might be the deeper problem: we can't admit the 2005 choice was wrong, so we're just slow-motioning a different model on top of it.