Fast Net News, DSL Prime, DOCSIS Report

"$5-8B in broadband stimulus money can create jobs building networks if spent wisely, and I strongly support it." I wrote early in January, when "everyone" was talking about a “broadband stimulus” in the tens of billions, My research had convinced me the carriers couldn't ramp fiber builds enough in less than 2 or 3 years to make a big difference in jobs. The first priority of the stimulus was jobs in 6-12 months. In the last 24 hours, I've received news of 800 layoffs at Time Warner, 11,000 at Intel, and 5,000 Ericsson. That's a good reason to emphasize jobs this year, which was widely report as the first directive.

I came remarkably close to the current proposal, but that wasn't because I had inside information. It was easy to guess that unserved rural areas would be put first. When I looked at specifics, it became clear that wireless could create more jobs quickly, and reach almost everyone with decent if relatively slow service. When I learned the time frame was one year, not three to ten, I knew the big fiber builds were irrelevant. It took Verizon four years to ramp to 3M homes/year. Starting with Verizon's figure of $700/home passed, you'd need to fibering 10's of millions of homes. That would be wonderful - and may be the right long term strategy - but it won't make many jobs in 2009.

Starting with that, the result was logical. Mostly rural, wireless first because it's faster to deploy almost everywhere. Only a limited amount for upgrades to fiber elsewhere. I came up with a figure based on company data and expert estimates of how much they could increase their build. None of the D.C. pundits (except Derek Turner) had bothered to ask the carriers how much they could build, so wound up recommending four or five times as much as the carriers could sensibly spend. Some of the numbers floating around D.C. would give the Bells about $3B even if they didn't add a single home or job to their current schedule.

I had recommended giving the big telcos a subsidy (perhaps 25%) on just about everything they would build beyond their current capex plans (FIOS 3M, U-Verse 6-7M) but that didn't get in. Most of that was diverted to rural areas. My guess is that's more generous than can be sensibly spent this year, but I'm not an expert on rural wireline.

$5-8B in broadband stimulus money can create jobs building networks if spent wisely, and I strongly support it. I strongly oppose ripping off the public treasury for the $20-30B the lobbyists want.

• Ubiquitious 3G/4G wireless to the 6-10M homes not covered by towers. Towers and backhaul in un-served areas are probably incremental and almost all U.S. job creating. This will allow the President to say in 2 years that 98-99% of U.S. homes can receive megabit service.

• Enough subsidy to encourage the telcos and cablecos to build everything practical. 60% increases in FIOS and U-Verse, DOCSIS 3.0 to 80-90% of the country.

• Unserved rural areas, any practical fiber or cable build significantly faster than wireless. Few projects are ready to go, unfortunately.

• Perhaps direct subsidies to poor families to get connected. This is a good thing and currently part of the program, but realistically will create few jobs. Adding more people to existing networks requires little labor and mostly imported parts. If public money subsidizes 5 M connections, I hope we don't pay retail. The direct cost of broadband in the U.S. is $5-10/month, meaning service at $20-50 is very profitable. Jules Genachowski, the new FCC chief, is a dealmaker. I hope he gets the price down to $13 for the low end and $20 for the faster service. The carriers still will get a great earnings boost.

• Perhaps community technology and training, which would almost all go to jobs.

  Bush eliminated the program for political reasons and less than expected results. Now that we need jobs, it's back on the agenda. I hope the folks writing the regs realize that the similar "demand stimulation" is a crock and a nationwide map can be made for under $10M. Almost everything beyond that - the beggars in business suits want $30-40B - would be wasted or go directly to corporate profits. Some proposals create few jobs:  DSLAMs, cable gear, and almost everything in wireless except towers need little labor and many of the parts come from abroad. A consensus I believe emerged that other available funds should help poor families directly with a subsidy similar to lifeline. [note added later – direct subsidies, such as half the broadband cost for families with kids in school under a certain income, did not make the final bill from the house.]

• Some very courageous people in D.C. are fighting to not waste public money. Their opposition includes some of the most powerful companies in the world, who are throwing everything they have in the hopes of more stimulus money. With luck, some of the excess will be diverted to subsidizing good connections for poor families instead of stockholders. [so far, the results look good]

• The Obama administration decided the absolute priority is to get jobs in 6 months, which means many of the efforts that make policy sense have no chance in the stimulus. More competition, nationwide fiber, etc. are good ideas that were ruled out because they wouldn't happen in 6 months. Some symbolic gesture towards net neutrality will probably be included to silence critics, but it's unnecessary. The key net neutrality battles in the U.S. have been won, although we need Free Press and others to make sure things don't slip in the details.

Hereﾒs Where itﾒs Smart to Spend Money

Today, 93% of U.S. homes can get broadband from cable companies, almost all with downstream speeds in the megabits. Only 4-6% of U.S. homes are limited to (painfully slow) satellite. Money meticulously targeted to that 4% is sensible public spending. I'd also support loans and small grants where that would make a difference. Direct money for broadband/computers to low and middle class families or similar (community technology centers) has some merit. Instead, the most discussed use of the billions will provide tax credits to large companies that would be building as much without the subsidy. Detroit may need a bailout, but AT&T expects to earn $10B this year after taxes. Beyond subsidies, I recommend providing nearly unlimited un-subsidized loans for 3G/4G wireless.  Some will default but will almost surely cost far less than subsidies of equal effect.

Ubiquitious 3G/4G wireless to homes not covered by towers.

$2-4B to get megabits of wireless to almost all of the currently un-served homes makes sense. Towers and backhaul in un-served areas are probably incremental and almost all U.S. job creating. Nearly anything else in wireless (upgrading to 3G/4G) is mostly non-incremental because it's related to the number of customers to serve. It's success-based spending that will happen anyway. But few rural towers are in the pipeline and they reach can reach unserved customers. (data reliability: I'm working from a detailed CTIA study from Costquest I've double checked with wireless experts.)

Enough subsidy to encourage the telcos and cablecos to build everything practical.

Verizon is set for 3M/year for FIOS, and would be doing well to increase that 50-60%, to 4.5-5M homes/year. Their cost is $700/home, and that build should be encouraged. God job creator, excellent network for customers. AT&T U-verse is at 16-17M end of 2008, and per their latest Annual Report is set for 30M in 2010, 6 to 7 million more each year. I've based my estimates on a 50% increase in U-Verse, which is ambitous given they missed their 2008 target. Many people don't believe public money should go to a build that's only 1 meg upstream, 10 meg down and creates few jobs because of the low labor demand, but I've included it. [To my amazement, the House bill includes even less, putting almost all the money into rural areas. The Bells actually serve the majority of “rural” lines, but that's something different.]

Unserved Rural Areas

4-6% of U.S. homes can't get anything except satellite. About half of them can be offered 50 megabits for under $400 each with existing cable lines. 1% are difficult to serve with a wire. I personally believe that last 1% should be offered (upgraded) satellite rather than spending $30K to connect some of them. That leaves 2-3% who will require new builds of either DSL or fiber.  Generous but not wasteful subsidies should make sense here. [I missed something here. Half that 4-6% can get cable video but not data, most small and older systems. Those can be upgraded for the $400 (old analog) or less. For the other 2-3%, I should have included new cable builds as well as DSL and fiber.]

Research and Mapping

A small amount of money, under $100 million, can be effectively spent on telecom research and availability mapping. Either could be ready to go in 3 months and will have a large return. However the data is overwhelming that “demand stimulation” does not work well.

---------------------

I also recommend nearly unlimited loans with very little subsidy to companies that need them. I would set strict limits on the 25% subsidy to what is clearly incremental builds that create more jobs than the companies already plan. As far as I can figure, giving more than $1B or so total to Verizon and AT&T combined is ripping off the taxpayers for the benefit of their shareholders.