Capitol Weekly runs a column in which Daniel Ballon suggests that the Sacramento network would cost $550m to build: Readers of this site know that I am pretty dubious that the Sacramento network will ever be built by the consortium that won the bid. Still. Please. Wi-Fi networks are estimated to cost around $150,000 per square mile. These numbers are well known. You add more nodes and costs go up. If you look at the now-well-received Minneapolis, Minn., network, US Internet now estimates $24m (up from $20m) to put 45 nodes per square mile. The 55 sq mi city will cost a whopping $440,000 per square mile to build, although that was supposed to include some fiber buildout (those details are sketchy in the documents I've found).
So, gentle readers, where does Ballon get the $550m figure? By looking at the cost to build the Sacramento airport and extrapolating its cost for its limited area by the city's dimensions. Ballon, as is disclosed in his linked bio but not on this page, works for the Pacific Research Institute, which is a think-tank that receives funding from Verizon and SBC, but that wouldn't be salient to disclose here, would it? (Ballon's background is in very small things, by the way: molecular biology and biochemistry, not telecom policy.) PRI has links to Big Tobacco, and ties to the Heartland Institute, which I have extensively covered in years past.
Now the issue is not that I disagree with Ballon's conclusion that a Sacramento network might cost vastly more than predicted. The original estimate doesn't contain enough nodes, and more than doubling the number from the 18 to 20 nodes planned (as reported in the Sacramento Bee 5-Nov-2007, and not refuted as far as I can tell) to 40 to 45 nodes would increase costs. They wouldn't double, because nodes are just a part of the overall cost of the network. But it's more likely a $15m to $20m network than a $7m to $9m one. (An anonymous commenter tried to tell me that the 18 to 20 nodes per sq mi figure was incorrect, but didn't reply to a request for the source of their information.)
Rather, the point is that Ballon has ties to beholden interests. It's fascinating that he mentions an existing competitive fiber provider in Sacramento with such positive praise--I never heard of SureWest, but he says they have 30 percent of the region's market, although not which market. I checked SureWest's site, and they have 30 percent of about 200,000 homes they pass--about half the households in the city. That's significant. And there's no comparison between fiber and Wi-Fi. The availability of Wi-Fi is in no way a challenge to the voice, video, and data triple-play and triple-threat that SureWest offers.
Ballon wants to paint the Metro Connect Sacramento network as government subsidized because the municipality may--but has not committed to, to my knowledge--shift some services from one set of private companies to another set of private companies. I thought that was the point of competition?
Glenn,
Thank you for your thoughtful comments.
As I write in the second paragraph, the actual cost of building the proposed 1Mbps network "would require...doubling the proposed $7-9 million investment."
You write above, "But it�s more likely a $15m to $20m network than a $7m to $9m one."
I'm not sure that I see the disagreement here. You say that costs would increase but "they wouldn't double." Isn't $15-20m more than double $7-9m?
The $550m number is based not on building a 1Mbps network, but rather a 4-6Mbps network. I apologize that "faster, fee-based subscription service" did not articulate this point clearly enough. The comparison of a 1Mbps network to the much faster airport network was made in the Sac Bee article. I merely extrapolated what a similar network could cost if it covered the entire city. I don't believe you challenged the math (but for the record, Terminal A = 275,000 square feet and Terminal B = 216,000 square feet. The city of Sacramento is 5,110 times larger than the size of the airport, making the potential cost $562,000,000).
Instead of challenging the research, you decided to attack the researcher. I prefer to debate the facts, so I will not respond except to correct a key mistake. In the interest of full disclosure, PRI received no money from Verizon or AT&T (or any other telco) in 2007. Donations in previous years were made before I joined the organization.
I will also pose this question regarding your last point: If the plan is merely to "shift some services from one set of private companies to another set of private companies," why can't SMC survive without an anchor tenancy agreement? Why not obtain the necessary permits from the city, build their network, and then compete on equal ground with all other available commercial products?
Again, I greatly appreciate your taking the time to comment on my piece.
All the best,
Daniel
I appreciate your response. The article wasn't clear on that point, and thus I have to agree with you that the cost could easily double to provide the basic tier. But Toronto has demonstrated that you can achieve extremely high throughput without spending what it requires for dense, interior, high-availability, multi-purpose Wi-Fi. The airport's costs aren't comparable to outdoor networks for a variety of reasons, including the cost of working in security-restricted spaces. I would believe that a 4 Mbps to 6 Mbps network might cost over $30 million but not over $500m.
I appreciate also your forthrightness in disclosing that no telcos have made recent contributions. Because of the general lack of transparency by thinktanks of all stripes about their corporate funding, it is often difficult to know whether opinions are bought and paid for.
(I get to claim purity because my reporting has been part of what has, in fact, led to a general decline in ad revenue in the field in which I write about. Every time I write about how municipal networks aren't working out, I am indirectly but palpably removing money from my pocket.)
I don't see anchor tenancy as that strange a situation compared to lock-in deals that Verizon et al. have required of municipalities for years for infrastructure investments, especially where wireless is involved. In many cases, municipalities are exiting multi-year agreements with incumbent providers and competitive providers to shift services to new operators, still private.
I had another thought here: The reason that the airport was so expensive is that they are building a WLAN not a WWAN. A wireless local area network (WLAN) in an airport can run at 20 to 30 Mbps of throughput with thousands of simultaneous users across acres. Gigabit Ethernet backhaul aggregated traffic. Uptime availability has to be at five 9s (99.999%) because many airports are using Wi-Fi for operational purposes. (Sacramento isn't, but they still have the WLAN parameters in general.)
A wireless wide area network (WWAN) has much lower requirements. Top speeds aren't guaranteed. 1 Mbps is the target, whether for cell or Wi-Fi WWANs. Backhaul is slower, with wired connections (at cell towers), fiber, or wireless links aggregating dispersed transmitters.