The Writing is on the Wire

This headline kinda says it all: Cablevision is Adding 3,200 Consumer VoIP Lines per Week in New York .

If voice over IP telephony is really growing that fast, then the old phone network's days are truly numbered.

Comments

[rich]

I dunno...we'll still need something to get DSL over, won't we? I don't see it as ending so much as evolving.

[garym]

Shrinking, yes. Vanishing, hardly. If nothing else, there are lots of people in rural areas who don't have high-speed Internet access or don't choose to deal with the cable company.

And the greater competition is coming from cell phones.

[s.m. koppelman]

Or rather, since the incremental cost of providing phone service in densely populated areas once the buildout has been paid for is low, VoIP will drive traditional telcos' rates way down. A lot of telcos' traffic is already VoIP, and by the same token much of that VoIP traffic is flowing throuigh the telcos' grids, which are paid for by phone service.

The problem is that in rural and remote areas, the cost of providing traditional phone service is much higher and might not exist at all without government-mandated universal service subsidized by markup on the profitable service in dense areas.

Rural and remote areas also don't have digital cable TV or DSL and they won't any time soon. Their 500-channel TV needs can be met by much more cost-effective consumer satellite service, but can satellite broadband internet carry adequate VoIP, or does it have too much latency?

The question is how to best continue to provide affordable universal phone service while still freeing up telcos to compete effectively with VoIP providers who don't have to share the costs of providing universal POTS. Unless we're going to stop subsidizing rural phone and elecctrical service in order to get some people to abandon the countryside (which might not be a bad idea), my vote would be for a tax on VoIP service that receives inbound traffic from the telephone grid.. that is, VoIP with a phone number. Proceeds would go into geographic pools that subsidize the cost of providing POTS to low-density, high-cost-per-customer areas. As cheap, effective, low-latency broadband internet comes to those areas, whether by the telcos themselves, or cable TV, wireless or powerline broadband, the subsidy would be removed and the tax would be reduced.

I'm sure someone who actually knows a thing or two about these issues could come up with something much better.

[Highway]

I think there are just going to have to be areas that won't get 'cheap, effective, low-latency broadband'. How much are people in high-density areas to be expected to subsidize those who, for the most part, choose to live way far away?

Actually, I'm sure that someday there will be a high-speed internet system that can transmit over long distances wirelessly, be it satellite or terrestrial. But until then, I don't think 'universal broadband' is really a worthwhile goal. Yes, people miss out on information on the internet, but if they really want to read Reason.com, they've gotta make the choice.

Additionally, how expensive is POTS service to provide to rural locations after the initial infrastructure is in? I realize maintenance costs are higher, but how much of the overall costs of service are really scaled up? That kind of information needs to be known before we make decisions about what to subsidize.

[s.m. koppelman]

See, there's the rub. If POTS is to survive, it has to be able to compete on price with VoIP. Because the telcos are providing service at uniform rates everywhere from cities to rural areas, that option isn't really open to them. Get rid of universal service or free up pricing and you'd quickly see city folk and suburbanites paying $4 a month for unlimited domestic calling with all sorts of bells and whistles, and folks in the middle of nowhere would all get ham radio licenses and rig up a gateway to one or two shared phone lines at the nearest general store or pay hundreds of dollars a month for basic phone or cellular service, carrying the full cost of the buildout and maintenance in their area.

New suburban development probably wouldn't get POTS at all in that scenario. My guess is you would probably see one company per new subdivision running fiber for broadband internet, or none at all if the power company offers internet. Residents would choose between cheap cellphone service and maybe satellite TV or cheap VoIP bundled with cable TV or electricty.

[Andrew D. Todd]

The physical plant of the the telephone network is basically durable. It is physically in the ground, and there it stays, barring extravagant effort to remove it. Telephone rates do not protect the network, so much as they protect the interests of the people who provided the money to build the network in the first place, or rather, the interests of the persons who have since bought out those persons. These people did so on the assurance that the telephone company had a franchise, and that the stock was essentially as safe as government securities. As a matter of political reality, they will have to be paid off on some terms or other. The plain sense of the agreement was that they were buying into a government-sanctioned monopoly.

The critical part of the telephone company's plant is in the form of ""true subscriber loop," that is, the last five hundred feet or less, before the individual line is plugged into a thick ganged cable containing many wires. Once you have a hundred subscribers hooked together, you are most of the way home, because the economies of scale are beginning to work in your favor. The going rate on E-bay for a switch machine capable of providing DSL service and flipping telephone traffic into VOIP is about fifty dollars a line. This is of course for a machine designed to work in a central office, which has not been weatherized. If you want to use the machine close to the subscriber, you may have to build some kind of shed for it to live in. Once you have your hundred subscribers aggregated, connecting them to the outside world is comparatively simple. You can use a microwave relay, or an optical fiber, according to circumstances.

For rural areas, the technology of choice is the microwave relay, not the landline. In rural open spaces, interference between microwave relays is comparatively unlikely. If you were to follow the old Rural Electrification Administration practice, the rule would be that in exchange for being connected into the network, each subscriber would allow his property to be used to connect other people further out. Given the rate of depreciation of electronics, it is probably reasonable to ask each subscriber to pay a substantial initiation fee, on the order of a couple of thousand dollars, and take out a loan if necessary. You simply cannot apply thirty-year amortization to equipment which is obsolescing at a moore's law rate.

Here is a link to a small cooperative in Colorado, with a good deal of information on economics, technical details, etc. As near as I can make out, they have been fighting a losing battle with Qwest over the right-of-way need to pull their couple of dozen subscribers together.

http://www.rric.net./

[Wireless Networks]

We are moving into a time where last-mile networking technologies will explode with choice.

It's not just POTS vs. Cable vs. DSL. Now you have a plethora of wireless networking technologies which operate on licensed and unlicensed spectrum.

MMDS and LMDS are giving way to WiMAX. Proprietary wireless technologies are exploding. 802.11b is moving rapidly from a wireless LAN technology to an ultra low-cost wireless WAN technology. Satellite is still making a play, although it's uplink technical challenges are enormous.

We're going from a data-over-voice network (modems on POTS lines) to a voice-over-data network (VoIP), but the real future will be anything-over-whatever.

So, buy some 802.11b gear and some good antenna', start your own broadband ISP, and sell voice service to your customer base. It's a Libertarian dream world.

[M. Simon]

Suppose nationwide the build out is going at the rate of 50,000 per week. That is 2.5 million a year.

If there are 50 million lines to connect that is going to take 20 years at the current rate. Time enough to work things out if we don't panic.