9 ms·
Here's a similar situation that turned out much better than expected. Back when I grew up in Texas in the 80s, the power utilities were regulated monopolies (l
by diafygi 12y ago
Here's a similar situation that turned out much better than expected.
Back when I grew up in Texas in the 80s, the power utilities were regulated monopolies (like pretty much everywhere else). They owned both the power plants and the wires going down the street. Then in the 90s, the utility lobby convinced the state that it would be much better for "competition" if utilities were "deregulated". The state went along with it, split utilities in to "wires" and "production", and deregulated production[1] (i.e. let anyone build a power plant and sell power at any price).
This was great for the current utilities because they were the only ones with existing power plants, so they could collude and jack up the rates[2]. However, in the early 2000s oil companies/prospectors started realizing that all this natural gas they were fracking could be used in power plants, and they could be the ones who owned the power plants.
So there was a huge boom[3] of gas power plants in the 2000s as many non-utilities built and competed on the wholesale market to provide power. This drove the price in the late 2000s and early 2010s down significantly since gas was so cheap. Deregulation actually worked in eventually driving down the price due to competition!
The difference between Texas utilities vs. FCC/ISPs is that the grid operators were still regulated and had to let anyone who sold power to use their grid. This effectively created the equivalent to the "Title 2" common carrier for the electric grid. The competition is fierce on the power production side, and the parts of the old utilities that were split off to compete are starting to go bankrupt[4].
So if the FCC wants competition, they need to follow the Texas system and split the cable company into regulated "wires" and deregulated "ISP" parts. Unfortunately, the existing monopolies have seen what happened in Texas, so I don't think they'll push for deregulation.
[1]: http://www.ercot.com/about/profile/history/ http://www.ercot.com/about/profile/history/
[2]: https://en.wikipedia.org/wiki/Deregulation_of_the_Texas_electricity_market#Electricity_prices https://en.wikipedia.org/wiki/Deregulation_of_the_Texas_elec...
[3]: http://www.poweracrosstexas.org/wp-content/uploads/2013/01/PAT-2013-Issue-Brief-Electric-Generation-Mix1.pdf http://www.poweracrosstexas.org/wp-content/uploads/2013/01/P...
[4]: http://www.nbcdfw.com/news/business/TXU-Oncor-Owners-Vote-for-Bankruptcy-257126791.html http://www.nbcdfw.com/news/business/TXU-Oncor-Owners-Vote-fo...
- lbotos 12y agoSo in this model, you are saying we should have companies that are responsible for just laying literal wires to connect houses to some type of neighborhood switch and then from the switch side on is "ISP" territory and they are responsible to connect to the "greater internet"? I'm just wondering how the physical roles would be split and literally where.
- mason55 12y agoThe reason for the split is that you don't want to have a bunch of companies trying to lay wire everywhere. It's very easy to build 10 power plants to connect to the grid vs. trying to run 10 sets of power lines everywhere. The problem is that internet wire technology moves so much faster than power wire tech. Even in just the last 20 years we've gone from copper phone lines to cable to fiber. As long as the last-mile tech continues to improve it will be hard to work this scenario because what incentive does the "wires" company have to re-do all their infrastructure from fiber to the next big thing?
- jessaustin 12y agoEven in just the last 20 years we've gone from copper phone lines to cable to fiber. Is there any reason to suspect that today's fiber (albeit possibly "lit" by tomorrow's fiber modems) will be insufficient twenty years from now? Are we approaching some sort of theoretical limit to the number of beams of light that one glass fiber can transmit?
- AnthonyMouse 12y ago> Is there any reason to suspect that today's fiber (albeit possibly "lit" by tomorrow's fiber modems) will be insufficient twenty years from now? Are we approaching some sort of theoretical limit to the number of beams of light that one glass fiber can transmit? No. And the premise that we've gone from twisted pair to coax to fiber in a period of 20 years is similarly silly. Telephone wire is essentially telegraph wire and has been in use since before the American civil war. Coax is from the same century and had been in use for cable TV before the internet was even a thing. Even modern fiber optic cable is older than, say, TCP/IP. The only reason fiber wasn't used to begin with is that the other cable was already in the ground before the internet existed.
- marcosdumay 12y agoI get the impression that the equivalent to that would be for the FCC enforce that ISPs can not discriminate based on content origin, and content creators must be left unregulated. Or, in other words, net neutrality. The idea that one entity owns the wires, and rent to another that connects the wires into the network is madness. Both entities will fight to no end, and bad service will always be caused by "the other one".
- Alphasite_ 12y agoExcept that it actually works just fine, see BT OpenReach in the UK.
- SkyMarshal 12y agoWhether we get Title II or not, the last mile needs true competition. YC Pres Sam Altman has a good blog post on it: http://blog.samaltman.com/net-neutrality http://blog.samaltman.com/net-neutrality
- techsupporter 12y agoI disagree with your assertion that deregulation has been good for the Texas power market. (I lived through that same transition and have been participated in it for most of my adult life.) For one, still-regulated cooperative and municipal power utilities still have lower rates than fixed-term "competitive" utilities. That said, I wholeheartedly agree with the idea of splitting data delivery into poles/wires and content parts. This is where I want municipal broadband to go. Have the county or the city or what-have-you own the physical plant and run it like the city water system (yes, I know, Texas has private water systems; don't get me started on Monarch Utilities). Any competitive provider that wants to hook up to the system and provide services would be welcome to do so.
- ams6110 12y agoThe problem is that cities and municipalities are poor stewards for these resources. They use the maintenance of the infrastructure as black holes for patronage employment. Had a water line break (on the city utility side of the meter) and the city turned up to fix it with three trucks and six men, for a job that a private plumbing contractor and excavator would have needed one or at most two workers to complete.
- zanny 12y agoIf the Internet were more mesh friendly this would make sense, but as long as you need to pipe all the data in a town through a backbone provider someone holds way too much power in the relationship. With power companies third parties can run their juice over lines that goes to homes without touching another plant. The state can maintain the switching stations, but otherwise power never needs to pass through another third party. Or in a state like PA, you may never get power from your choice provider, because you get to pick your power company here. So I can buy from some business out of Erie who is only serving power locally while I get all my juice from another company, but state regulation says that I get metered billing from my service and not power provider. That can't really work with the Internet. Fundamentally, if the state owned the wires and the switching stations, they would own all the infrastructure. There is no equivalent to power plants in IP - unless you want private switching stations, which means you are not emulating the electrical grid (usually). It is kind of brain dead simple for a municipality to provide a dumb wire and routing, though. I think it is worth noting that where I live (a town in PA called Kutztown) the borough a decade ago started a township owned telecommunications business subsidized by taxpayers. It operates as a semi-private entity, and since the huge township investment (it is still paying off the 15 million investment) the data rates have never gone up once, and the pricing has only gotten worse - it started at 10 / 10 mbps for $25, and today the rates are 10 / 1 for $30 and 10 / 8 for $60. The company has no incentive to improve their infrastructure and bureaucracy guaranteed the deployment was hugely inefficient for taxpayers to foot the bill. The only benefit was that the traditional telecom giant in the region, Service Electric, dramatically slashed prices in the borough in response to the town sponsored option - but their service now stays just as expensive as HomeTown Utilicom, and neither improve service or reduce pricing because now it is just a duopoly where one only expands through taxpayer money and the other has more lucrative markets to leech dry where they are the only cable provider. So what I am saying is that the public option may not be the sunshine and rainbows everyone dreams of. The state is dramatically inefficient at building anything. They will spend ten times the money and get a tenth the return because they have only tertiary cost pressure and massive bureaucratic bloat. The perfect scenario is where you can somehow get infrastructure at private enterprise prices (and that means you had to get a profit motivated cost deferring entity to initially deploy the infrastructure to keep the expenses low) while somehow having the state pay off the business, and then give away wire access to its constituents for free or cheap. And then you have to constantly figure out a way to apply market pressure to improve the infrastructure, rather than just stagnate on the first attempt.