> The danger with government nationalizing telecommunications networks is that once it is done, innovation and quality go way down.
Who needs nationalization? The market for transit is reasonably competitive. Local municipalities could install fiber along the roads they already maintain at a modest incremental cost, then offer the service to residents for a monthly fee to pay off the bonds without even spending any taxpayer money.
All you really need from the federal government is to have them do something about incumbent ISPs actively interfering with municipalities that want to do that.
> 4G/5G are decent options
No they're not. The nature of wireless is that it's cheaper if you have a low population density, because one tower for hundreds of people is much cheaper than installing hundreds of miles of fiber for hundreds of people.
It flips completely the other way in anything resembling a city. To get a fraction of the bandwidth available from fiber, you'd need a tower on every street corner, which isn't dramatically less expensive than installing fiber (especially when you count all the spectrum you have to pay for) and is still slower even then.
And cellular is even less attractive when you have Starlink -- then you don't even need the towers. It's great for rural areas. But it's hardly going to have enough aggregate bandwidth to let all of New York City watch Netflix in 4K.
Starlink does require towers because the data has to be beamed back to the ground. The more bandwidth you serve in an area, the more towers you need. However, in areas with low bandwidth demands where cellular towers are currently as dense as they are in order to reach every part of the area, Starlink does make sense.
Starlink, Cellular and ground based internet serve different points on the density /mobility curve.
> Starlink does require towers because the data has to be beamed back to the ground. The more bandwidth you serve in an area, the more towers you need.
That's a ground station, not towers. The "towers" are the satellites. And the number of grounds stations you need doesn't really have much to do with how much bandwidth you use, because they can use directional wireless or laser communications at higher power levels than ordinary devices, so they're not the bottleneck.
The reason to have a large number of ground stations is so that there is one within range no matter where a satellite is. But they have plans to build an inter-satellite mesh that would remove that requirement as well:
In principle if they wanted to serve a higher density area they would need more satellites so they could each cover less area, but that's not really its purpose. Dense areas can more than justify a fiber network. What the satellites get you is world-wide coverage from one network, particularly including the places that don't have existing coverage because their density is too low.
And the combination of that doesn't leave much left for cellular to do. You have a very high bandwidth fiber connection at home, at work, backing the wifi in any kind of a hotel or coffee house, and then Starlink for when you're in a rural area or in a car and you want to look at a map but aren't near any of those. What's left for cellular?
Probably fit on your car though, which then gives you a hotspot for your phone wherever you are.
I suppose if you're out in the middle of nowhere away from both any building with fiber and any vehicle with Starlink then cellular might be useful, both those are also the kind of places without any cell reception.
Currently not every building's WiFi is accessible to you. Second, WiFi only has a limited range. If your house is at the edge of settlement for example and you take a walk in nature (or just the nearest park!), you won't get any WiFi after the first dozen meters. That area may very well be covered by cellular though. Cellular has a much higher range and will still be very important.
Who needs nationalization? The market for transit is reasonably competitive. Local municipalities could install fiber along the roads they already maintain at a modest incremental cost, then offer the service to residents for a monthly fee to pay off the bonds without even spending any taxpayer money.
All you really need from the federal government is to have them do something about incumbent ISPs actively interfering with municipalities that want to do that.
> 4G/5G are decent options
No they're not. The nature of wireless is that it's cheaper if you have a low population density, because one tower for hundreds of people is much cheaper than installing hundreds of miles of fiber for hundreds of people.
It flips completely the other way in anything resembling a city. To get a fraction of the bandwidth available from fiber, you'd need a tower on every street corner, which isn't dramatically less expensive than installing fiber (especially when you count all the spectrum you have to pay for) and is still slower even then.
And cellular is even less attractive when you have Starlink -- then you don't even need the towers. It's great for rural areas. But it's hardly going to have enough aggregate bandwidth to let all of New York City watch Netflix in 4K.