| Stimulus: Questions to Ask To Save $5B |
| Saturday, 03 January 2009 14:06 |
|
$20B in broadband stimulus is being allocated in the next week, and I believe some of the decisionmakers have been fed misleading information. The public comments of both consumer and corporate advocates are loaded with claims that contradict what I've reported from the network designers. These questions I believe can help find the key issues and consider whether you have the facts you need. The answers are my opinion and research, but I urge you to check with technical experts you trust as well. Always happy to provide data and sources (including those who disagree) for a journalist or policy person. Should stimulus money be primarily used for job creation or is it appropriate to use a major share to improve the companies' balance sheets and dividends? Does this create jobs in six months or even two years? "How much will each company build without subsidy? How much with a subsidy of perhaps 25% or generous loans? What's the maximum possible given the time it takes to ramp up new projects? How much of the subsidy will go to investments already in the pipeline? Do the details in the build undercut the main goals of creating jobs and improving U.S. networks? Getting good answes to those questions is one way to tell whether the bill is on target and how much waste. Should stimulus money be primarily used for job creation or is it appropriate to use a major share to improve the companies' Balance Sheet and dividends? Obama's answer is ..., and nearly everyone other than company lobbyists immediately says the money should go to jobs. However, the public is providing hundreds of $billions to the balance sheets of AIG, Fannie Mae, and all the big banks, so sometimes that's the right move. This is a key question, $5-8B is enough to provide 25% subsidies to just about every sensible project practical in the U.S. in 2009 and 2010. $20-30B total Does this create jobs in six months or even two years? Several current proposals would create few U.S. jobs, the priority. Others, like building wireless towers and fiber to the home, are mostly construction labor and equipment built in the states. The likely stimulus items that lead to few jobs are any additions to existing cable, telco, and wireless networks that need little construction. Adding a new DSL or cable customer in mostly adding equipment genreally not made in the U.S. U-Verse "Fiber to the Node" has a fancy name, but is really DSLAMs moved to field cabinets. It was carefully designed not to require much construction or labor, and much of the equipment used is from abroad. Similarly, upgrading existing wireless towers to 3G/4G is mostly equipment, much from abroad, and therefore adds few U.S. jobs. Providing connections and computers to those who can't afford them I think a good thing, and a far better use of stimulus dollars than many of the lobbyist's proposals that wind up in waste or company profits. Because most of those connections can be served by networks in place, they add few jobs. The cost is mostly equipment from abroad. I believe the same is true of computers, but that's not my expertise. The "indirectly created" jobs take much longer to develop and almost certainly are far fewer than the "100,000 jobs for every $5 billion" the CWA is claiming. That can not be true if the money is not spent on more broadband or wasted.
My analysis is that AT&T and Verizon can increase the ten million lines they are doing a year by 50-70% in two years
Ask ten people "How much will you build without subsidy? How much will you build with a subsidy of perhaps 25% or generous loans?" If they give you an honest answer, you are very close to knowing whether the money is well spent. 4 companies - AT&T, Comcast, Verizon, and Time Warner are about 60% 0f U.S. lines. Ten companies are over 80% of lines. Most of the smaller companies will follow similar patterns, and no others are likely to produce milions of lines in two or three years, much less six months. I asked those questions, and used the companies' financial statements as well as expert opinion to make estimates when they didn't give a direct answer. I found the maximum likely usable spending over two years is $5-8B at a 25% subsidy. Some companies are upgrading 100% in two years without a subsidy (Comcast and Cablevision are unsung heroes.) Verizon and AT&T are 80% of U.S. telco lines. They are already adding 10M lines/year, and could probably raise that to 15M with help. 3G/4G wireless towers reach 93% of more of the U.S. population and few new ones are built. That's the largest single opportunity, with tower construction ready to begin in 90-120 days widely if the government helps. More at. How much of the subsidy will go to investments already in the pipeline? Communications companies are so large they do and much have plans 3 and 5 years out, and generally follow them in normal times. Billions of public dollars mean this is not normal times and increases up to 50-60% may be possible and a natural stimulus. The figures I've heard the large carriers are asking are so high they are asking for money for already planned and budgeted work or total waste. Is there a sensible alternative, ready to go, that doesn't waste public money? The key question is whether to deploy DOCSIS 3.0 cable, telco DSL (including U-Verse,) or fiber to the home, like Verizon FIOS, UTOPIA, and Burlington. DOCSIS 3.0 has deployed to 15M U.S. homes in the last six months and will be at 50M+ by 2010. It delivers 50 megabit service 95+% of the time and probably better (160/120 shared.) It can be done more quickly than the telco alternative; Comcast is doing 50M homes by 2010. Fiber (designed for 200 meg and reliability is best but costs $700/home (Verizon) or more (smaller companies can't be as efficient.) DOCSIS at 50 meg is about $100/home, and still under $400 to most (not all) of the extreme rurals. It turns out the decision for 2009 and 2010 is easy to see; there's plenty in the stimulus to build all the fiber, DOCSIS and U-Verse DSL the industry is capable of installing. working thoughts:
|