I was troubled to see Broadband Communities publish an odd and misleading story about Palo Alto in the May-June issue [pdf]. Authored by Stephen Blum of Tellus Venture Associates, a consultant that has been hired by Palo Alto in the past, it showed a remarkable level of ignorance about community owned fiber networks and broadband more generally.
The title alone, "Can FTTP Work in Palo Alto?" is just odd. Why exactly would FTTP not work in Palo Alto? It works in hundreds of other cities and towns, most of whom are less well positioned than Palo Alto for such a venture. A more honest title would have been "Consultant Argues Never Used Financing Mechanism Also Won't Work in Palo Alto." Blum made a very good case for that narrow argument but fails to lay out any convincing evidence that a variety of other models are doomed.
Parts of the article can only be called cable and DSL boosterism - such as repeating the talking point that AT&T's U-Verse and Comcast already offer "high levels of service at competitive rates." Competitive to what? Neither can deliver the speeds offered by modern fiber networks and are only "competitive" if one ignores the much slower upstream speeds, higher prices, lesser reliability, problems of oversubscription, and poor customer service one gets from those providers.
Reminds me of "Slick Sam" from Lafayette and the "functional equivalence" between DSL and FTTH.
Blum apparently knows better - that Palo Alto residents are "happy" with the existing services because they have not spontaneously marched down El Camino Real demanding faster speeds at lower prices. This is the wrong measure - reminiscent of the now oft-quoted Henry Ford line that if he asked people what they wanted, they would have said "faster horses."
The number of specific errors in this piece are many, and have been well documented by those familiar with the history of Palo Alto's studies. I want to focus on just a few. Let's start here:
Overall, 79 percent of households would have to pay $3,000 apiece to fully fund FTTP construction costs.
YIKES! Cue the foreboding music! Palo Alto has something like 25,000 households. If 20,000 of them paid $3,000 then the City would have $60 million in addition to its present $14 million dark fiber reserve - a staggering $74 million of theoretical money that has nothing to do with anything. I know of no network that has been built in this manner.
This is an absurd measure for whether a network is feasible. Networks are not financed in this way, partially because, as the author adroitly notes, it doesn't appear likely to work. Community owned networks are financed using a few common methods, most often revenue bonds issued by the utility. Palo Alto's past studies of this approach reflected a desire to avoid that path and the results of those studies in no way determine whether a city owned FTTH network is feasible in 2013 given the present assets and environment.
The user-financed model remains a peculiarity and quite possibly will have a role to play in the future (though almost certainly not to finance the entirety of a system). Palo Alto would be crazy to hinge its decision of whether to invest solely on the feasibility of each home owner paying its full connection cost up front.
In examining the likelihood of success for Palo Alto, it makes sense to consider similar communities that have made the investment:
Although there were some apparent FTTP successes (for example Bristol, Va., and Cedar Falls, Iowa), cities that had more in common with Palo Alto, such as Alameda and Provo, Utah, were failing.
There is nothing "apparently" successful about Cedar Falls or Bristol. They are unambiguously stunning successes, with take rates north of 70 percent and have led to thousands of jobs. And both can deliver a gigabit anywhere in town at a moments notice at rates a fraction of what major carriers charge. But he believes Alameda (with an older HFC cable system) and Provo (having to deal with strict state laws not present in California) are more relevant comparisons. There are some 140 other citywide networks that might be more relevant, but Blum ignores them.
He ultimately concludes that Google's experience in Provo will somehow inform local government decisions around network investments. This is some of the worst advice I have read. First of all, Google's costs are different than any other firm, let alone a local government because it already runs one of the largest fiber networks on the planet (possibly the largest). The most brilliant engineers on the planet work for Google. It doesn't publish its costs and is a private sector firm with far different motivations and incentives than a local government.
In short, there might not be a worse comparison than Google for a local government evaluating its own plan for meeting long term telecommunications needs.
All of this being said, Palo Alto could rationally choose not to invest in a FTTH network. It would have to compete against Comcast and AT&T, who engage in predatory tactics while federal regulators ignore potential Sherman Anti-Trust violations.
Given the many wonderful aspects of the community, particularly for people who don't like winter, maybe DSL and Comcast cable will be good enough for the heart of Silicon Valley. That is their choice, not mine (my wife and I love Minnesota winter). I just hate to see such an imbalanced and inaccurate case made suggesting it could not work.