In Tuesday’s election voters approved the muni broadband opt-out I discussed last time by a whopping margin of 493 votes out of  43,913 or one percent. This was a strange election because virtually none of the voters knew what they were voting for.

There’s a good record of voter perceptions on hyper-local websites like Nextdoor and Patch where average voters ask each other about ballot questions and the like.

The overwhelming takeaway from mining these sites is that voters believed the city had to opt-out of the state’s ban on municipal broadband networks before the council could even discuss programs to get poor people online, provide broadband assistance to non-profits, or even to provide it to city departments. So the question of the city obtaining voter approval to create a $400 million network devolved into permission to hold meetings.

What Voters Said

Here’s a sample of the comments on my neighborhood’s Nextdoor group.

I attended a Town Hall meeeting where a retiring city council member said that the City needed voters approval for exemption from the State law to even begin a discussion about the feasibility of studying this issue. In other words “We the people need to give them the permission to discuss the benefits or not of municipal internet services.” I’m voting “yes” for them to study this issue because private industry does not serve everyone and not everyone can afford commercial internet fees.

This was a meeting held in one of the city’s five council districts, known as “wards” where two council members addressed voters. Here’s another one echoing the same line:

The ballot is simply allowing the city to look into options, so anyone who tries to scare you with “government taking your choices away!” is literally telling you the opposite of the truth.

That said, the goal (in my opinion) should be to nationalize internet access as a part of our national infrastructure.

And some people believed that the city needed to act to preserve home rule rights under its charter:

At the Ward 4 meeting on Oct. 12, our two Councilors made it sound like 2G is to retain Lakewood’s Home Rule rights against legislation passed by the state in (I think) 2015. In my brain I think of it as a place holder. It is not authorizing the start of any Internet service provided by the City. That would also have to be approved by the voters. One person asked a very pertinent question: “Why do we have to do this now?” I don’t recall hearing about any deadlines to retain our Home Rule rights. So I think it may be possible for this question to arise again in the future if there is serious talk about Lakewood doing its own Internet. Unless I’m missing some key points, I don’t care if it passes or not. I hope others who were at the meeting jump in with what they heard.

Another commenter responded to her with the “right to talk” idea:

You are correct, [redacted]. This issue would authorize the city to even TALK about ever doing city Internet service. There is no particular deadline to do this authorization, but nothing would ever proceed unless the authorization is passed. [Councilman] Dave Skilling mentioned that a great number of Colorado towns/cities have passed a similar authorization, but of them all, only two cities proceeded to set up Internet service.

A somewhat fanatical member with an IT background echoed the misconception again:

My understanding is that local government cannot spend any money researching municipal broadband without a vote to opt out of Senate Bill 152. Over 100 Colorado cities have opted out of this bill and have gained the right to explore whether municipal broadband is feasible. Lakewood’s city website does mention Lakewood currently has no plans to build out it’s own broadband network. I think it’s worth looking into.

Our alternative is Comcast. Again, this is a vote to investigate, not to build.

So the voters were not generally thinking they were voting to authorize the construction of a network and still the measure passed by the thinnest of margins.

Contradicting the Standard Story

The standard narrative about muni broadband in Colorado created by advocates is that dozens of cities have taken back their authority and are busily building their own networks. Here’s a story from 2015:

Staggering Approval

The landslide victory was no surprise. Last year, nine communities asked voters the same issue of whether or not they wanted the ability to make local telecommunications decisions. That right was taken away 10 years ago by SB 152. Two other communities took up the question earlier this year with 75 percent and 92 percent of voters supporting local telecommunications authority.

A few larger communities, such as Boulder, Montrose, and Centennial, presented the issue to the voters and reclaimed local authority in prior years. This year, most of the voting took place in smaller, rural communities where incumbents have little incentive to invest in network upgrades.

This year, results were similar as the majority of voters supported local measures with over 70 percent of ballots cast. In Durango, over 90 percent of voters chose to opt out of restrictive SB 152; Telluride voters affirmed their commitment to local authority when over 93 percent of votes supported measure 2B. Many communities showed support in the mid- and upper- 80th percentile.

But it hasn’t really played out that way. The vast majority of Colorado cites that have passed SB 152 opt-outs (140 cities or counties) have not proceeded to build networks. At this point, only Longmont (pop. 86,000), Meeker/Rangely (pop. 4,450), and Fort Morgan (pop. 28,000) have city-wide municipal networks.

Construction or planning is underway in Fort Collins (pop. 170,000), Loveland (pop. 78,000), and Estes Park (pop. 6,339). Some cities have installed business service networks (Durango, Glenwood Springs), patches of dark fiber (Centennial), or pilot projects (Cortez), but the rush to pass ballot measures has not been followed up by significant network construction.

Boulder insists that it wants to build a network, but it recognizes that it can’t without taking over the electric service provided by Xcel. The power company has refused two offers from Boulder and the city is now trying to seize the power system through legal means. This isn’t going well, but the day may come when Boulder builds a hemp-fired power plant and a solar broadband network.

Confronting Reality

The real story is that plenty of cities and counties have added SB 152 opt-out questions to regular election ballots in order to have some degree of “local control” over potential fiber assets. This is an easy sell for city councils chafing under legislative restriction, but the consumer rebellion that might lead to large-scale network construction just isn’t happening.

The only cities of any size that have gone this way can build networks cheaply because they operate the local power and water companies and have invested in fiber to deploy smart grids and smart meters for at least a decade. They also have oddball populations: Fort Collins is a college town that looks up to Boulder, and Longmont is a Boulder suburb.

Boulder is basically Berkeley with snow, a college town with a low vaccination rate, a ban on biotech farming, and a massive fear of gluten. The most popular major at the University of Colorado Boulder is “Integrative Physiology”, a B.A. program on “the role of physical activity in human health and function across the lifespan and the responses of different organisms to various forms of stress.”

Both Longmont and Fort Collins are well served by Comcast and CenturyLink, with speeds up to a gigabit over cable and major fiber build-outs underway by CenturyLink. Longmont has slashed prices for fiber service to $55 until Dec. 31 to stem customer losses to Comcast’s triple- and quad-play service bundles and CenturyLink’s $65 gigabit fiber.

The Rural Story Makes More Sense

The major success stories for muni broadband are small towns (fewer than 10,000 people) with no cable service and 5 Mbps DSL. This is the story in Rio Blanco County and in the better known town of Ammon, Idaho, the suburb of Idaho Springs that created the so-called “Ammon Model” of network financing.

The small towns benefit from regional middle-mile fiber projects and regional Internet exchanges. People tend to overlook the fact that Internet service isn’t strictly a local problem; you can wire up an entire town easily enough, but the Internet is a global network.

So transit plays a big role in rural Internet regardless of who provides it. In Colorado, our ISPs connect to the Internet in Dallas, Chicago, and Salt Lake City. In Idaho, networks connect in Salt Lake City and in Seattle. The cost of transit establishes a lower limit on consumer prices.

An Aside on the Ammon Model

While not directly related to Lakewood’s issue, the Ammon Model of muni broadband financing was raised by one of our city council members following the lobbyist presentation in support of the ballot measure. The lobbyist, former Arvada, Colorado mayor Ken Fellman, confessed to never having heard of it despite the large amount of play it garners in muni circles.

Ammon, Idaho is a suburb of Idaho Springs with a population of 15,000 on seven square miles and a very rapid growth rate: between 2000 and 2010, it grew by 125%. In the late ’90s, Ammon pulled its own fiber to connect a couple of government buildings after calculating that it could do the job a lot cheaper than the contractors that answered its RFQ.

This led to a practice of interconnecting its municipal water system, parks, and other government buildings. As of 2015, Ammon had no cable service and balky 5 Mbps DSL. When the phone company refused to upgrade its speeds, the city once again asked for bids on a modern broadband network and found it would be cheaper to build than to buy.

The Most Conservative Financing Model Ever, With a Twist

After Ammon created a Local Improvement District and purchased a boring machine for cable ducts, it devised a scheme to have consumers cover construction costs though a commitment to pay $3,300 per home for an Ethernet hookup. When it found the interest was strong enough, it floated a bond measure to finance this fee over twenty years for roughly $17 a month per home.

Here’s the clever part: instead of installing dark fiber or trying to play ISP, Ammon created a software-defined network with network slicing, just like a modern 5G network. This set it apart from all the other muni fiber networks in the US (but not the world).

The SDN allows each subscriber to reach a switching center that provides basic lifeline support and a menu of services. The menu includes residential broadband and business broadband (differentiated between business hour vs. 24/7 support) at a range of speeds from 15 Mbps to 10 Gbps.

The city doesn’t connect consumers to the Internet, but it will connect them to services that do as well as services with a more local flavor. While consumers save money by using the city’s home-grown cable pulling service, they lose it on the transit fees paid by small ISPs.

Operational Details

The customer pays the city for installation and operation of the network, including repair and maintenance; they pay their ISP for Internet access. The typical bill is $17 for installation (as noted,) another $16 for operation, and $10 – 100 for the the actual service, not counting fees and taxes.

You can see the menu by clicking this link and logging in as coademo/coademo. Customers can easily switch ISPs and can even subscribe to multiple service grades if they want. Ammon Fiber is not at all neutral, but it is highly functional.

The main drawbacks of this model are that, like Google Fiber, it’s slow to install and it prioritizes rich neighborhoods over poor ones. Because the city does all the installation with its own help and its own equipment, would-be customers have to persuade their neighbors to sign up before anyone gets service.

The city won’t act until it ensures a 55% take rate by legally binding customer pledges and anyone who doesn’t like the service is still on the hook for the $3,300 fee whether they use the network or not. These conditions obviously won’t work in many communities, but they work in Ammon according to this six hour video from the inaugural event. This looks like a reasonable way to provide broadband in an underserved community even if it’s not a generally acceptable approach.

Conclusions

Contrary to popular myth, Colorado is not witnessing a taxpayer revolt against commercial broadband. A number of cities and counties has seized the authority from the state to build broadband networks through regular election ballot measures, but only a half dozen have proceeded to build anything other than a free Wi-Fi network or a business-only network.

Muni broadband is a viable option for underserved small suburbs and exurbs, but it’s not especially attractive in larger towns. Economies of scale in providing world-wide connectivity work against small providers even when taxpayer subsidies cover startup costs. This is reflected in the fact that half of Colorado’s six municipal broadband networks share a common backhaul plant.

The regulatory status of ISPs providing service over public dark fiber and SDN networks is intriguing, but I suspect they’re strong candidates for Title I treatment (like dial-up ISPs) even if overall FCC doctrine changes back to the Title II policy we saw in the days of the Wheeler FCC. But we’ll leave that question for another post.