A proposal to stimulate broadband deployment

by: John Windhausen

Fri Jul 27, 2007 at 14:48


Mr. Windhausen will be liveblogging on OpenLeft this evening at 7 PM EDT.

Many congratulations to Sen. Durbin for instigating this "outside-the-box" discussion of broadband policies and universal service.  Broadband services are more than simply an evolution of our existing telephone technology and services; broadband networks are and will fundamentally alter the way our economy functions, the way students learn, the way workers are trained, the way medical care is delivered, the way communities share information, and the way government governs.  So we need to think big about our broadband policies and not fall into the trap of only trying to "fix" existing programs.  I suggest that we start thinking of our overall objectives and develop a real broadband plan.

(More in 'there's more'...)

John Windhausen :: A proposal to stimulate broadband deployment
First, let's establish that our goal should be broadband deployment to 100% of American homes.  And let's assume that by broadband we mean more than the current generation of technologies (such as DSL and wireless 3G), which are becoming unsatisfactory because of the increasing demand for video (which uses more bandwidth than web browsing or e-mail). 

Second, let's acknowledge that broadband deployment is expensive.  Using "back-of-the envelope" calculations, let's assume it costs about $2,000 per home to deploy fiber to the home (Verizon is doing this for somewhat less per home, but its deployment is not in rural areas).  With 115 million households in the U.S., it will cost about $230 billion to deploy fiber to every home.  This is too much money for the federal government to build these networks on its own.  We will need to rely upon the industry to fund the bulk of this expense, while using government to "prime" the market. 

To encourage greater industry investment, we need to consider policies that combine rewards for the private sector for investing while also ensuring that the broadband pipes remain open and affordable.  This is a delicate balance, but both are necessary.  The private sector can be given accelerated depreciation, investment tax credits, and matching grants (such as in Idaho) to promote investment.  But a big broadband pipe is useless if you can't use it as you wish.  So, the companies receiving these benefits must serve the public -- they must permit entrepreneurs and and service providers to have access to the network.  Our European colleagues have leaped ahead of the U.S. in broadband deployment by adopting an open regime.  They have shown that greater usage stimulates greater investment. 

The federal government can also seed the market with funding of its own to supplement spending by the private sector.  Unfortunately, the existing government programs are not all that well-equipped for this purpose.  The Broadband Loan program administered by the Rural Utility Service is hampered because by the fact that it primarily consists of loans, not grants.  The application process is exceedingly difficult, largely because applications must demonstrate enough financial strength to be able to pay the loan back.  Those who can get the loans may be the companies who need them least.  On the other hand, the federal Universal Service Fund is about $7 billion, and about $4 billion of this goes to subsidize high-cost (mainly rural) areas.  While $4 billion may seem like a lot of money, it's a pittance compared to what we need to attain the goal of universal broadband.  Furthermore, some of this $4 billion goes simply to maintaining the existing copper telephone network and is not devoted to broadband deployment.  While some advocates are focused on "fixing" or "reforming" the existing universal service program to focus more on broadband, consider whether that is worth the monetary and political capital.  The existing programs are replete with arcane rules and entrenched interests that will fight hard to retain them. 

Rather, it may be smarter to create a new program focused solely on broadband deployment.  Recipients of the funding must demonstrate that they will build networks providing at least 100 Mbps to the home and preference would be given to rural areas.  A $10 billion program per year could be funded from general revenues, from excess spectrum auction revenues, or from a $1.50 monthly fee on every phone bill. Finally, since the program would focus on the up-front costs of deployment, not annual maintenance, the program could sunset once its mission is accomplished. 

This proposal may or may not be feasible, but perhaps it will provoke even better ideas geared to make broadband a reality.

John Windhausen, Jr. has worked in the communications policy arena in Washington, D.C. for over 20 years. He served as senior counsel to Senator Hollings, Senator Inouye, and the Democrats on the U.S. Senate Commerce Committee. He participated in the drafting and passage of the 1992 Cable Act, the 1993 Spectrum Auction Act, and the Telecommunications Act of 1996. He also served as President of the Association for Local Telecommunications Services (ALTS), where he helped lead the facilities-based local telecom industry in upholding the pro-competitive principles embodied in the Telecommunications Act. He recently started Telepoly.com, a new informational web site on communications policies.


Tags: , , , , , (All Tags)
Print Friendly View Send As Email
Good ideas, John (0.00 / 0)
John,

I like where you're going with this, and have a few comments.

I agree its best to create a new program rather than try to mix this with either the RUS program and the copper-focused Universal Service Fund, with its entrenched interests and political quicksand-like dynamics.

I think the only "safe" way of doing this (in terms of insuring open access) would be to require a fully wholesale model, similar to the UTOPIA model discussed in another post by Paul Morris...sort of the fiber optic version of what Google and others are proposing for the 700 MHz band. Having government funds to help finance the cost of network deployment would presumably encourage more companies to pursue the resulting opportunities associated with provision of retail services on a wholesale fiber platform.  My understanding is that this has been an issue in the past, though I'm guessing it is becoming less so over time, a question that Paul Morris might be able to address based on his experience with UTOPIA.

I have some concerns about letting incumbent network operators be eligible for this within their service areas, though there are also some real advantages to getting them involved.  My concerns would relate to potential conflicts of interest, in that they'd be simultaneously building out next-generation fiber networks based on a wholesale model, while operating current-generation networks offering retail services. 

On the other hand, it could be a genuine win-win opportunity for incumbents to make the migration from their legacy network and business model to a future-proof network based on an Internet-friendly open-access wholesale model

But overall, incumbents' track record for good faith is not a great one, so you'd probably want to think through the potential conflicts of interest and their implications before opening the government subsidies to incumbents.

Your $10 billion per year number sounds reasonable in terms of its contribution to build-out costs.  If we assume $250 billion in total and network deployment over 5 years, this public funding would pay for 20% of the total network costs.

Do you have any more specific thoughts on how the "rural preference" you mention could be implemented?

I think the network capacity requirement needs to include both downstream and upstream capabilities.  Though the cable industry's HFC networks are pretty capable and a very sensible evolution of historical cable TV networks, they are especially weak with regard to upstream capability, so I'd hate to see government subsidies going toward expansion of these networks.  The same is true of any DSL-based network that spends money to extend copper to customers' homes, even if they are able to deliver 100 Mbps.  If you're going to spend money to lay copper (twisted pair or coax), it won't cost that much more to install fiber, since labor is the biggest cost component.  And in very rural areas, it might make most sense to extend fiber to some point, then use wireless links for the final customer connections, which would be a lot cheaper and more flexible than installing copper access links.


Sorry about the bolding.. (0.00 / 0)
Sorry about all the bold text.  I'm not sure what happened there.

More "back of the napkin" calculations. (0.00 / 0)
Sir,

While I would be the last person to suggest that FTTH wouldn't be a great asset to the US, I believe your "back of the napkin" calculations might need to be revisited. If we are willing to accept the numbers you are citing based on Verizon's experience, we also need to understand that Verizon is in the process of selling off Maine, New Hampshire and Vermont because I believe Verizon understands that the cost per home would be significantly higher in these regions.

One thing that was excluded that would skew your numbers is that there is no allowance for all buildings that are not residences, most of which would require connections. Please let me also point out that there are great expanses of the rural US that would need to have extremely long lengths of backbone put in place, adding to the average cost per home, making it well above the $2K per home average cost figure you are using.

To put this is perspective, I will use Vermont, as it is a state that I have a working knowledge of, even if my data is somewhat out of date.

As I recall (please keep in mind, I am older now and my memory isn't what it used to be) Vermont has something like 89,000 miles of paved roads with an unknown but probably equal number of dirt roads, many of those inhabited. Assuming we use a reasonable cost of $10K per mile for fiber (hung, not buried) times the 89,000 miles of paved roads we can easily calculate that the paved roads alone will cost $890,000,000 and if we assume that there are half as many dirt roads that need fiber deployed we find we have $1,335,000,000!

We also know that Vermont has a population of 635,000 people (plus or minus) and it is my understanding that Vermont is made up of 240,634 households. If we assign a value of $1,000 per household for installation, including all parts and labor (this is not too far out of line based on Verizon's recent admission that their average for installing Fios is 9 hours per home) we can estimate a cost (for only this portion of the deployment) of $240,634,000 which when added to the $1,335,000,000 comes to a total of $1,575,634.000. This would bring just the cost of installing fiber to roughly $6,550/household without any allowance for central office buildings, equipment etc. In my opinion, I would suggest to you that there is never going to be a return on investment in this specific case.

Now, I have no way to figure out if Vermont is an average that we could expand on for the entire US, but if it were, it is my understanding that there are 105,539,122 households in the US (total) and if we apply the calculated Vermont average cost across the entire US, we find that the total cost for the fiber being connected to every home (but not including all the supporting infrastructure) works out to be $691,281,249,100 and I cannot even begin to estimate the time that would be required to build out such a project.

And with all of that done, one good ice storm could cause tens of millions of dollars damage that would take weeks to repair.

In addition, fixed internet access is a great thing but mobile access is also something that would need to be added into your calculations.

In summary, I believe the best answer to this problem is going to be a combination of technologies.


Good points Ken (0.00 / 0)
Ken,
Your comments underscore the value of a strategy that optimizes network design based on  real world conditions, notably density. 

My preferred buildout strategy is actually to extend fiber as far as seems economical, and then to extend wireless the rest of the way.  Then, as demand increases, fiber could be incrementally extended over time, to more and more premises (as you note, not just homes, but businesses, govt buildings, etc.)

The wireless portion would always have value, even if fiber was later extended to all premises in an area, since it would provide the added benefit of mobility, and perhaps also some valuable network redundancy in case of fiber cuts, storms, etc.

The value of the wireless piece is why I keep pushing for opening up the broadcast white space for unlicensed and/or "public" use. This spectrum has strong propagation characteristics and therefore could be very cost-effective for long distance links in rural areas.  And having that spectrum available would also make it easier to deliver higher bit rates than if we just relied on existing WiFi spectrum bands.


[ Parent ]
Donate to Open Left









QUICK HITS

Friends of the Earth thanks the OpenLeft community for the ideas you generate and your contributions to the progressive movement.


blog advertising is good for you
blog advertising is good for you
SEARCH

   

Advanced Search