Wi-Fi: can’t even give it away

Not only is Earthlink getting out of municipal Wi-Fi, but it can’t even give it away:

The Atlanta-based Internet service provider said Tuesday that it could not find a buyer for the $17 million network and that talks to give it to either the city or a nonprofit organization had failed.

City officials have said it would cost taxpayers millions each year to operate the network.

“It's been an unfortunate situation,” Chief Executive Officer Rolla Huff told The Associated Press. “It was a great idea a few years ago, ... but it's an idea that simply didn't make it.”

EarthLink, which will give current customers until June 12 to switch to another provider, said it even offered to donate the Wi-Fi equipment to someone and give them an additional $1 million.

Earthlink’s announcement implies that their other business are sound, but, frankly, if they don’t grow in wireless access it seems that they are doomed to be a declining seller of dialup access and reseller of broadband that competes with their cable and DSL suppliers.

Still, Earthlink will probably outlive the imminent death of the municipal Wi-Fi movement, which has been facing imminent death for more than a year. With Earthlink abandoning systems already built, other cities have decided not to even start.

It must be a pretty lousy business that you can’t even make a business with a zero capital cost? At least the new Iridium was able to turn a profit after acquiring its assets for a half cent on the dollar. At least satellite phones have a small niche of customers who have no alternative, as opposed to connectivity in a big city where there are cable, DSL, dialup, local hotspots, libraries and increasingly 3G mobile phones.

However, the local nonprofit, Wireless Philadelphia, seems to be in denial:

Wireless Philadelphia and the City of Philadelphia continue to work together to ensure a positive future for Philadelphia's municipal wireless network and nationally-recognized Digital Inclusion program, the vision of which is to provide all citizens with access to essential technological resources for education, employment, and other life opportunities.

Good intentions and noble causes don’t obviate economic reality. It’s good to see such “frill” services held to some measure of economic accountability, even if the core functions of government are not.

Municipal Wi-Fi: stick a fork in it!

Last month Earthlink admitted that its municipal Wi-Fi business was a mistake and it was bailing out. As I noted in my posting “End to the Wi-Fi mirage”:

This is probably the beginning of the end of attempts to build self-supporting municipal Wi-Fi systems (as opposed to those subsidized like parks and libraries as a “public good”).

This weekend, even the NYT has noticed (wistfully) that the city-subsidized plans are also on their way to oblivion. Not surprisingly, the system design overestimated the hotspot range and thus underestimated the number of base stations required. (WiMax partisans: take note.)

The obvious lesson is if public-private partnerships seemed like a risk-free way for government to get something for nothing, in fact if the private entity can’t make a profitable business, the partnership is doomed.

The municipal Wi-Fi fanatics are hoping to deny economic realities and have someone subsidize their pt approach to social action, urban renewal, etc. etc. But it’s time to stick a fork in the whole movement — it’s done for good. If people really care about making Internet accessible — rather than spending lots of money or building cool toys — then libraries and community centers with free Internet access are a lot less expensive way to (mostly) accomplish the same goal.

End to the Wi-Fi mirage

Earthlink is now trying to sell or close its municipal Wi-Fi operations — perhaps because it lost $80 million on municipal Wi-Fi last year (versus $20 million in 2006). Apparently some of the failed services are getting turned off. Earthlink also pulled back from the ill-fated Helio MVNO venture with SK Telecom.

This is probably the beginning of the end of attempts to build self-supporting municipal Wi-Fi systems (as opposed to those subsidized like parks and libraries as a “public good”).

The cause is not that different than Ricochet’s failure in the 1990s. Yes, Ricochet had slow nonstandard modems while Wi-Fi is ubiquitous and cheap. However, the economics of building infrastructure coverage are the same, and today the desirability and adoption of substitutes (other access methods) are probably worse.

Most seriously of all, it seems like the era of paid Wi-Fi is heading towards extinction — because free Wi-Fi at restaurants and coffee shops is becoming more the norm. I had lunch in Mountain View on Sunday and it seemed every bar on Murphy Street had free Wi-Fi. Starbucks announced Monday that it’s offering (limited) free Wi-Fi — to match all the other free sites — further fueling the commoditization of Wi-Fi hotspots. About the only place that people will pay for Wi-Fi is in an airport, because you can’t easily go down the street to find a better alternatives.

Without crunching the numbers, my intuition is that WiMax is going to face the same problem. Yes, the radios have longer range, but you still have to build more cells (and negotiate access and install backhaul) for lots of cells. If WiMax fails, then Sprint's 4G strategy will fail with it.

Municipal Wi-Fi: what’s the point?

There were a lot of interesting sessions at the LA GMR. As with most conferences, I learn the most when the knowledge gradient is greatest (as long as I understand the subject well enough to know what’s going on).

So the first aha! moment at the conference was at the session entitled “WIMAX/Muni-WiFi Experiences: LA, Philadelphia.” Moderator Richard Grimes (of Wireless Capital Partners) introduced some of the common issues: the use of municipal Wi-Fi to provide universal access (to address the so-called “digital divide”), the ownership of the system (public, private, hybrid), and the expected impact of WiMax. (Here I focus on systems being built, and will consider the WiMax vaporware in another blog entry).

The first paper was by a friend of a friend (previous acquaintance) who by the end of the conference had become a good friend. Youngjin Yoo of Temple U. described his fascinating Philadelphia story. (The effect was accentuated by his use of Larry Lessig-style slides: 2-5 words per slide, black & white.)

Yoo briefly reviewed the history. In 2003, the CIO for Philadelphia convinced his political masters that it should be “the first major city in the US to declare we have a 21st century infrastructure.” (That they believed this suggests that the politicians are pretty dumb, but more on that later).

So in 2004, the city announces its plans, and forms the Wireless Philadelphia public-private consortium. The next year, the city plans to spend $15-18 million to build the network, until Earthlink proposes to run it as a business. Due to city delays, the pilot test doesn’t begin until the end of 2006. Today, there is limited coverage (which I’ll test in August) but the whole system is supposed to be done at the end of 2007.

The consortium used four “pillars” to justify its involvement:

• digital inclusion for four key political constituencies: low income families, single parents, senior citizens and HIV/AIDS victims. Yoo said that Philadelphia is one of the poorest cities in the US (next to the poorest place in the US: Camden, NJ), but the inclusion of the other groups suggested that some served in those groups would be non-poor (middle class or even affluent).
• tourism.
• small business development.
• e-government (although it was not clear what this meant).

The plan took off because of the combination of the socio-political (politicians trying to look good) with the techno-economic forces (Intel, Motorola trying to sell hardware), in what Yoo called “a perfect storm.” He and his co-authors mapped out the relationship of the different actors, as shown below:

At the end of the day, the plan had two major flaws. First, the purported beneficiaries were really stick figures to justify what people wanted to do. When asked, they gave many reasons why they wouldn’t use or benefit from the plan, as in this quote: “Why wasn’t the [budget for the initiative] spent on schools, or flue shots? Wireless is still a luxury and there are still many unmet basic needs.” Yoo said that for some non-profits, half of their constituents can’t even read but won’t admit it out of shame.

Second, Philadelphia is emblematic of the dubious technical and economic viability of muni Wi-Fi. The cost structure is pretty clear: one base station on a city lamp post every 1,000', or a 5x5 grid per square mile. Earthlink’s pricing plan was set early: $10/month for the underprivileged, $22 for everyone else. The presumed use of the network seemed to be residential and small business, rather than anything involving mobility (college students sitting in the park with their laptops).

But, as the entire session made clear, there is a very narrow window of demand for municipal Wi-Fi. If people are too poor, they don’t have the computers and computer skills to access a network (a related problem that all muni Wi-Fi advocates hope to solve someday). If people are too rich, they are also using other, possibly technically superior solutions. In Philadelphia, that’s Comcast cable modems and Verizon fiber optics (50 megabit) at $45/month. Even if promoters justify Wi-Fi expenditures based on a “public good” claim, they are obliged to give the honest cost, including any government operating subsidies.

Back to this “21st century” claim. 802.11b first caught on when Apple started including it in its laptops in 1999, 18 months (yes, 18 months) before the new millennium. 802.11g started shipping in December 2002. So is it worth spending $15 million to build an infrastructure that will be obsolete before 2010? And if it’s obsolete, where will you get the revenues to keep it operating? Still, the city of Los Angeles and many other cities keep marching forward with their plans.

All this prompted me to ask at the end of the session: If it’s not economically viable, why are we talking about municipal wireless at all? (Another academic said she plans to use this question to provoke discussion in one of her own studies).

One speaker mentioned the experiment of Lompoc, a small California military town. I looked up the numbers, and they would be laughable if not for the waste of taxpayer money. Here’s how the AP reported it in the Los Angeles Times:

A $3-million plan to blanket Lompoc, Calif., with a wireless Internet system promised a quantum leap for economic development: The remote community hit hard by cutbacks at nearby Vandenberg Air Force Base would join the 21st century with cheap and plentiful high-speed access.

Instead, nearly a year after its launch, Lompoc Net is limping along. The Central California city of 42,000, surrounded by rolling hills, wineries and flower fields located more than 17 miles from the nearest major highway, has only a few hundred subscribers.

That's far fewer than the 4,000 needed to start repaying loans from the city's utility coffers, potentially leaving smaller reserves to guard against electric rate increases.

Other cities (including Los Angeles) are proceeding ahead, with total spending by US cities in 2007 expected to exceed $400 million. A the AP noted, the consequences of failure carry a real price:

Without revenues they had counted on to offset that spending, elected officials might have to break promises or find money in already-tight budgets to subsidize the systems for the low-income families and city workers who depend on the access. Cities might end up running the systems if companies abandon networks they had built.

I love technology — and always want more — but sometimes the numbers just don’t work. Firms can make dumb bets and their investors pay the price — and thus investors reign in the most foolish ideas. However, when politicians do something stupid, they don’t pay the price — the public does.

I guess a public/private partnership can work out OK. The city cannot guarantee any shortfall, and should recover its out-of-pocket costs through a floor on the revenue sharing. Under these conditions, the firms face all the risks (and almost all of the upside) and so need to go into the deal with their eyes wide open.

Technorati Tags: LA GMR, Municipal Wi-Fi