> So if they just deliver 100/100 within a year or two, this is an epic win
It's unlikely that they (Starlink or someone else) will offer symmetrical speeds. It's not that they're looking to be mean to you. It's that uplink is harder and people use a lot more downlink.
Even if they dedicated as much wireless spectrum to downlink and uplink, uplink would likely be slower. We see this on traditional cell networks. They dedicate as much spectrum to uplink as downlink, but their cell tower is able to better transmit than your equipment and so the downlink becomes faster. Even with so many more users downloading than uploading (and causing congestion), the downlink is usually faster.
Newer wireless networks aren't going to dedicate equal spectrum to downlink and uplink (using time-division instead of frequency division to separate downlink and uplink). Again, this isn't to be mean to you. It's just a reality that people use a lot more downlink bandwidth than uplink. Starlink isn't immune from that reality.
Elon Musk has already said that they should be able to hit 500,000 customers, but that scaling to millions of customers will be difficult. They're going to have to cap how many people get service in an area and/or put in network management to make sure that some users don't use up all the bandwidth from others nearby.
They're also going to need to focus on downlink capacity to serve what users need. That doesn't mean unusable uplink. As you noted, 10-25Mbps uplink is an important improvement. But wireless internet options (including Starlink) will need to balance that with the downlink capacity users need.
> We could have paid $5,000 per house to lay it ourself but our own neighborhood couldn't come to consensus on that
Over 10 years, that's $42/mo. Over 20 years, $21/mo. That's non-trivial if the solution to your internet woes might just be a couple years away. It's probably one of the big reasons why wired companies won't want to be spending money expanding networks in suburban/rural areas. Let's say that you invest in a network expansion and you expect to make it back over the next 20 years. Then 3 years into your investment, Starlink, T-Mobile, and Verizon are all offering home internet service to your customers. Sure, your fiber network might be "better", but that will only attract some users. Others might get a package deal from their wireless carrier giving them a better price. Now you go from having 90% of households to 50% of households and the investment that you made probably won't work out. For most people, 100Mbps is plenty. Sure, some people love the low-ping times of fiber and love gigabit speeds. People like us here on HN. For most people, they want to be able to use Netflix/YouTube/Facebook/etc. and there's going to be a lot of competition for that market.
> Now imagine that at a national level
Realistically, this already exists on the national level in that we spend billions subsidizing rural connections. Starlink is receiving lots of government money to provide rural internet. I think a big question is whether Starlink is looking to grow well beyond what the government will subsidize and whether government subsidies will flow to other companies more. I'm sure that AT&T/T-Mobile/Verizon are all looking at what rural internet subsidies might come their way as they launch rural home internet.
We do have some political will to fix the situation, but it's a very expensive situation to fix for a lot of rural areas in a wired way. Should people in cities subsidize suburban/rural lifestyles? As a country, we pour money into roads, low fuel prices (even as climate change ravages the planet), rural telecommunications, etc. If every home in your area is on 2 acres of land, it's going to cost more to wire up the place, it's going to cost more to get roads everywhere, it's going to use more fuel to get from place to place.
We do spend a lot making rural internet happen. It's just an expensive proposition. Heck, Starlink is very expensive at a $550 startup cost + $100/mo. That isn't cheap competition to wired internet - and that's after large government subsidies that might end up being $2,000 per user. Starlink has received $900M in federal money and Elon Musk is hoping to serve 500,000 users so that would be $1,800 per user from the government. That's not the $5,000 your service provider wanted to extend fiber, but it's still a lot of money. Plus, Starlink is likely to be getting more federal money in the future (and they might end up serving a couple million users).
There is political will and we've spent incredible amounts of money over many decades and continue to spend even more. It's just hard to serve many rural areas. If one is in an area with 5 people per square mile, that's a lot of wire for very few users. Wireless/satellite might make the most sense since installing one thing could serve hundreds or thousands of users. Even 20-50 people per square mile can be a lot of work to wire up.
While wireless home internet is in its infancy right now, I'd expect it to get a lot better over the next 5 years. As you noted, your neighbor installed a 20-foot mount to get better reception. T-Mobile Home Internet customers are rigging up directional antennas mounted on the outside of their homes to get better speeds. Given that people install satellite dishes for TV, it seems very reasonable that we'll see wireless antennas installed to offer internet service. Again, when something is in its infancy, there are less options and it's less fully realized. But that will change over time.
I think the next 5 years will be an exciting time for home internet. I don't think that Starlink is going to be doing most of the exciting stuff and I don't think we'll see symmetrical connections, but I think we'll see great stuff that will bring better connections to people who need it and will bring competition to the marketplace to prevent monopoly providers from taking advantage of their customers.
My new neighbors are increasingly tech TLAs as the former generation sells their homes off for 2-3x what they paid 20+ years ago. Median house price in my neighborhood is ~$1.1M ATM.
> Applying this strategy to a micro data set from England between 1995 and 2010 we find a significantly positive effect, but diminishing returns to speed.
I think it is difficult to overestimate how much other rural internet providers have over-promised for their federal $$. I know that in my area (northwestern Montana), a variety of ISP's have received rural internet funds, and it seems like a fair bit of money has been spent on very little additional high-speed coverage. There are many towns from 350 - 2000 population that are under-served, even with a lot of federal subsidy (I'm told the local telco removed internet capacity from our town of 400 to provide it to a nearby school system. I know that if I pay $5K to get a telephone wire to my house, I can have dial-tone, but not internet, not because of distance, but because of capacity limits.) Perhaps when those ISP's applied for their grants, they simply underestimated the costs.
Regardless, we do have Starlink, and it is transformative.
Southwest Montana here: your situation sounds much worse than ours. I did have to build my own Wisp to get service to my home but in town there is decent HFC and fiber loop if you can afford it. I'm curious if you know for sure that you have regional backhaul capacity limits. I ask because I hear stories from local people here, including folks who should be better informed because they work in city planning, about limited connectivity. Their picture of the connectivity situation is wildly innacurate, in part because providers don't disclose the details of their installed plant.
My understanding is that the local telco (centurylink) would just as soon not be here. It certainly seems unwilling to make the investment necessary to provide more connectivity. Some smaller communities have limited fiber (Babb, St Mary) but there is none in East Glacier and Browning.
I see. So those towns are still microwave-fed? Around here I don't think we have any towns left that are microwave-fed. Gardiner had backhaul fiber installed a few years back. That was the last connectivity hole I'm aware of in this area.
It's unlikely that they (Starlink or someone else) will offer symmetrical speeds. It's not that they're looking to be mean to you. It's that uplink is harder and people use a lot more downlink.
Even if they dedicated as much wireless spectrum to downlink and uplink, uplink would likely be slower. We see this on traditional cell networks. They dedicate as much spectrum to uplink as downlink, but their cell tower is able to better transmit than your equipment and so the downlink becomes faster. Even with so many more users downloading than uploading (and causing congestion), the downlink is usually faster.
Newer wireless networks aren't going to dedicate equal spectrum to downlink and uplink (using time-division instead of frequency division to separate downlink and uplink). Again, this isn't to be mean to you. It's just a reality that people use a lot more downlink bandwidth than uplink. Starlink isn't immune from that reality.
Elon Musk has already said that they should be able to hit 500,000 customers, but that scaling to millions of customers will be difficult. They're going to have to cap how many people get service in an area and/or put in network management to make sure that some users don't use up all the bandwidth from others nearby.
They're also going to need to focus on downlink capacity to serve what users need. That doesn't mean unusable uplink. As you noted, 10-25Mbps uplink is an important improvement. But wireless internet options (including Starlink) will need to balance that with the downlink capacity users need.
> We could have paid $5,000 per house to lay it ourself but our own neighborhood couldn't come to consensus on that
Over 10 years, that's $42/mo. Over 20 years, $21/mo. That's non-trivial if the solution to your internet woes might just be a couple years away. It's probably one of the big reasons why wired companies won't want to be spending money expanding networks in suburban/rural areas. Let's say that you invest in a network expansion and you expect to make it back over the next 20 years. Then 3 years into your investment, Starlink, T-Mobile, and Verizon are all offering home internet service to your customers. Sure, your fiber network might be "better", but that will only attract some users. Others might get a package deal from their wireless carrier giving them a better price. Now you go from having 90% of households to 50% of households and the investment that you made probably won't work out. For most people, 100Mbps is plenty. Sure, some people love the low-ping times of fiber and love gigabit speeds. People like us here on HN. For most people, they want to be able to use Netflix/YouTube/Facebook/etc. and there's going to be a lot of competition for that market.
> Now imagine that at a national level
Realistically, this already exists on the national level in that we spend billions subsidizing rural connections. Starlink is receiving lots of government money to provide rural internet. I think a big question is whether Starlink is looking to grow well beyond what the government will subsidize and whether government subsidies will flow to other companies more. I'm sure that AT&T/T-Mobile/Verizon are all looking at what rural internet subsidies might come their way as they launch rural home internet.
We do have some political will to fix the situation, but it's a very expensive situation to fix for a lot of rural areas in a wired way. Should people in cities subsidize suburban/rural lifestyles? As a country, we pour money into roads, low fuel prices (even as climate change ravages the planet), rural telecommunications, etc. If every home in your area is on 2 acres of land, it's going to cost more to wire up the place, it's going to cost more to get roads everywhere, it's going to use more fuel to get from place to place.
We do spend a lot making rural internet happen. It's just an expensive proposition. Heck, Starlink is very expensive at a $550 startup cost + $100/mo. That isn't cheap competition to wired internet - and that's after large government subsidies that might end up being $2,000 per user. Starlink has received $900M in federal money and Elon Musk is hoping to serve 500,000 users so that would be $1,800 per user from the government. That's not the $5,000 your service provider wanted to extend fiber, but it's still a lot of money. Plus, Starlink is likely to be getting more federal money in the future (and they might end up serving a couple million users).
There is political will and we've spent incredible amounts of money over many decades and continue to spend even more. It's just hard to serve many rural areas. If one is in an area with 5 people per square mile, that's a lot of wire for very few users. Wireless/satellite might make the most sense since installing one thing could serve hundreds or thousands of users. Even 20-50 people per square mile can be a lot of work to wire up.
While wireless home internet is in its infancy right now, I'd expect it to get a lot better over the next 5 years. As you noted, your neighbor installed a 20-foot mount to get better reception. T-Mobile Home Internet customers are rigging up directional antennas mounted on the outside of their homes to get better speeds. Given that people install satellite dishes for TV, it seems very reasonable that we'll see wireless antennas installed to offer internet service. Again, when something is in its infancy, there are less options and it's less fully realized. But that will change over time.
I think the next 5 years will be an exciting time for home internet. I don't think that Starlink is going to be doing most of the exciting stuff and I don't think we'll see symmetrical connections, but I think we'll see great stuff that will bring better connections to people who need it and will bring competition to the marketplace to prevent monopoly providers from taking advantage of their customers.