Hacker News Reader

Comment by rayiner | Hacker News Reader

This article is a mishmash of ideas that don't really work together. The author advocates for treating telecom infrastructure as a public utility, but then uses AT&T as an example of an "intelligent" network that wasn't innovative. But the phone system, under the AT&T monopoly, was regulated as a public utility! It was a system that prioritized stability, universal access, and yes, neutrality, over innovation. That's the nature of public utilities.

The author suggests that the cable companies should be forced to lease out their hybrid fiber-coax networks to competitors. Leaving aside the fact that these are private networks built with private money, there is the issue that this would destroy any incentive on the part of the cable companies to enhance and upgrade that infrastructure. Why pour billions of dollars into upgrading a system only to get razor-thin returns?

View on HN

>The author advocates for treating telecom infrastructure like public utilities, but then says that competition is the answer. These are not compatible ideas!

Why not? I can choose to get my electricity and gas from any of about 6 different companies, and water from at least 4. They still come in through the same cables and pipes, but price and packaging varies.

>The author suggests that the cable companies should be forced to lease out their hybrid fiber-coax networks to competitors.

Again, here, BT have to do that as part of being able to have a (near-[1])monopoly on building those networks in the first place. That means people can get fibre to the cabinet/premises without having to pay for BT's overpriced and underreliable service, even if it runs over their fibre network for the last mile.

[1] BT have a near-monopoly nationwide, while some areas have their own local networks in addition, but these can not be nationwide. Virgin (formerly NTL/Telewest) have a parallel national network, which uses coax for the last mile more like US cable[2], but has a far lower overall availability, being in a general case only available in cities and major towns, and in many of those then only in certain areas of them.

[2] BT's network is fibre to the cabinet with copper from the cabinet to premises in most populated areas now(up to 80 mbit), with some lucky area having fibre to the premises(up to 1 gbit), and rural/undersubscribed areas having copper to the cabinet (generally, maximum rates of ~12mbit or worse; many can only get 0.5 mbit). Virgin's network can generally get 100mbit to most/all premises it serves, which is a far more limited subset of BT's. Local ones vary wildly and may be anything up to 10gbit, but are generally around 100mbit.

>>I can choose to get my electricity and gas from any of about 6 different companies, and water from at least 4.

Huh, seriously? I don't think I've ever lived anywhere that didn't have monopoly gas, electric, water and sewer providers. Monopoly trash pickup is not that uncommon either. Perhaps that's a UK vs. US thing.

It struck me as weird, too. I dunno. I've lived in or spent time pretty much all over the US in the last 10 years, and as far as I could tell, I've never had a choice in my local water provider. Gas and electric, usually not, though I suppose there are places where some choice is available.

A situation like that would probably induce legislation or regulation to prevent price gouging. There are a few areas where it happens, e.g. in Hull there is only one ISP, who are by all accounts terrible, but can afford to be as there is no competition.

With water, yes, it's heavily regulated as a public utility. Gas and electrical is a more interesting story. Here in California, we've had some adventures with deregulation of electricity (while still maintaining natural monopolies over provision). It was a complete clusterfuck.

Decoupling the lines from the power-generation works okay as long as you have excess power generation capacity. E.g. California-style deregulation worked fine in Illinois.

Power generation also has the problem that building new generation capacity is extremely expensive and takes years or decades. The generating companies in California used that to their advantage by shutting down quite a lot of existing generating capacity so that there would be a supply shortage that the market couldn't respond to quickly and that predictably caused prices to skyrocket.

It isn't likely that would be repeated in the market for interconnection because it's a lot faster and cheaper for a competitor to run a strand of fiber to a central office than to build a multi-GW power generating station.

Where I live I can pay a small premium to have my power generated by renewable sources. Electrons are fungible on the grid so I assume they have it worked out on the back-end of the provider just supplying X load to the grid.

Yep, it's a fairly complex (as in, I don't understand it fully, although I've read a little) interaction of subsidies, varying prices, and the same of standby capacity (i.e. not actual capacity, but the ability to ramp up to generate x GW in y seconds should it be needed, to cover spikes - generally by gas power stations, which are able to vary output power far more quickly than most types, and by pumped storage, which can release up to a certain amount of energy per day for very short, large spikes, for example one one caused by people putting the kettle on when popular TV shows finish).

Most of the trading is automated and humans aren't involved, as prices are being constantly renegotiated as demand changes, and supply may also be throttled down or up for fuel-consuming non-nuclear sources (nuclear runs as close to max output as they can due to running costs being consistent through output levels, and fuel costs being negligible in any case) to maximise profit - generally by increasing output as prices rise, but for wind and other renewables, sometimes output is artificially limited as a way of manipulating subsidies, which can actually be more profitable for the operators if less power is sold under the right grid conditions.

All the data is open, which lets some interesting sites appear, such as http://www.gridwatch.templar.co.uk/index.php . Mousing over the source numbers shows a brief overview, and the site also includes international purchase/sale data.

The BT situation is very different, though. BT started out as a government-owned corporation. When it was privatized, the government developed an incentive scheme to ensure that the company had sufficient returns on its investment to continue to upgrade its network, and wrote that into the prospectus.

Public-utility "last mile", far from being opposed to competition, would facilitate it. The local or state government would own the ROW and maintain the fiber - maybe thru a contractor whose role is restricted to that function. Then your household or business line goes to a box where different companies can compete for slots/ports/whatever, and then you would choose which provider to connect you from there to the big pipes.

The "to the premises" provider would make a regulated profit, just like utility-model water and power companies, and the upstream providers would pay the going rate for their connections to customers, and compete for customers with rates, policies and over-the-top services.

Having a "last mile" monopoly, either embodied in the municipal government or in a regulated utility like British Telecom, can work. However, ensuring that the monopoly has sufficient incentives to maintain and upgrade the infrastructure is complex. See this article which discusses the issues presented by the British Telecom privatization: http://www.oecd.org/sti/ieconomy/1909801.pdf.

While it's theoretically possible, it's not what the author actually suggests. He suggests forcing Comcast, TWC, etc, to simply lease out their networks wholesale to competitors. This destroys the incentive these companies have to continue to upgrade their networks. There is an obvious reason why the author advocates that approach instead of the one you suggest: state and local governments have no money, and are extremely inefficient at building infrastructure projects to begin with.

We recently moved back to using BT as an ISP with their Infinity 2 FTTC service and I have been very pleasantly surprised, both with the installation (which was very quick and painless) and the resulting service.

It also looks like we might get full FTTP in the next year or so which will be even better.

Worth noting that we had BT as an ISP years ago and they were awful so I was reluctant to switch back - but the seem to be making significant investments in their network and their customer services - as well as some pretty good add on offerings like their sports channels.

What was most interesting is that Sky, he had bought the ISP we used to use (Be), were desperate to keep our business - they offered to reduce the 20Mbs connection we had to £8 a month...

So from what I can tell, the regulation of the BT-owned "last mile" seems to have created a market with real competition - which can only be good for consumers.

> So from what I can tell, the regulation of the BT-owned "last mile" seems to have created a market with real competition - which can only be good for consumers.

You shouldn't conflate the "regulation" that net-neutrality proponents are calling for in the U.S. with the "regulation" of BT. BT started out as a public corporation, which was privatized in the late 1980's and early 1990's. This privatization was accomplished with an express eye to ensuring that BT would continue to have incentives to invest in its network, and investors had notice of the terms of the arrangement.

What net neutrality proponents in the U.S. are arguing for is an additional layer of restrictions on existing networks that were built with private capital.

Yes, I didn't think the UK style regulation would apply at all in the US - just pointing out that it does seem to have worked here, which I think we can all agree is actually fairly surprising. After all, I don't think the other privatised services - particularly energy, water and railways, have worked out quite as well for consumers.

> The author advocates for treating telecom infrastructure like public utilities, but then says that competition is the answer. These are not compatible ideas!

They are compatible if you decouple the infrastructure owner from the service providers, as it happens in other industries.

Reply on news.ycombinator.com