Evidence sheet · s06

Why Cities Should Build Their Own Broadband Nearly everything a person has to do now happens over an internet connection. School, work, doctor appointments, job applications, paying a water bill. When the connection is bad or too expensive, all of that gets harder, and it tends to get harder for the same households every time. Two researchers who have looked closely at this are Thorsdottir, writing in 2022, and Baptiste, writing a year later in 2023. Both of them study what happens when a city decides to build the network itself instead of waiting for somebody else to do it. After going through what both of them found, I think city governments should build and operate their own broadband networks.

Thorsdottir was writing in 2022, and the first thing she establishes is how uneven the access already was. She writes that "in the three counties surveyed, the number of households that could choose between two or more wired providers at speeds above one hundred megabits was smaller than the number that could choose none at all, a distribution that has been stable since the last buildout" (Thorsdottir 119). What Thorsdottir is saying here is that more households had zero options than had two or more options, and that this had not changed in years. She then explains why the situation stayed frozen: "A private firm answers to a return threshold before it answers to a service area. Where the density is thin, the threshold is not met, and the map simply stops" (Thorsdottir 121). In other words, companies stop building where they cannot hit their target return, so the map stops in the same place. The middle of her article is about a town she calls Kestrel Basin, which put in its own fiber after years of asking. Of that build she writes that "the city took nine months to pass the enabling ordinance and thirty-one months to light the last address, and the utility board absorbed the network into the same billing and repair structure that already served water and electricity" (Thorsdottir 124). This means the city did not have to invent a whole new department for it, it used the utility structure that was already sitting there. She adds that "the operating advantage was not technological. It was that the entity building the line had no incentive to leave any address on the line unserved" (Thorsdottir 125). What she means is that a city has to serve the whole city, so it does not skip the houses at the end of the road.

Figure 1. Broadband subscription rate by household income 41% <25k 58% 25-50k 72% 50-75k 84% 75-100k 91% 100k+ A year later, in 2023, Baptiste picks up the same question but looks at a large group of cities rather than one town. Baptiste writes that "of the forty-six municipal systems examined, thirty-eight were meeting or exceeding their debt service coverage projections by year five, and the eight that were not shared a single feature: they had been built as retail-only systems in markets where the anchor institutions had already contracted elsewhere" (Baptiste 269). So most of the systems were paying for themselves by the fifth year, and the ones that were not had one specific thing in common. Baptiste goes on to say that "the question worth asking is not whether municipal networks can succeed but under which structures they have, and the record on that is now long enough to be boring" (Baptiste 270). Baptiste is saying that we already know these networks can work and that the evidence for it is not new anymore.

The main objection is the money, that a city borrows tens of millions of dollars and the taxpayers are the ones stuck with it if the thing does not work. This comes up in Baptiste too. Baptiste writes that "the eight underperforming systems carried a combined shortfall that in three cases required transfers from the general fund, and residents in those cities are correct to describe the transfer as a cost they did not vote for" (Baptiste 274). So in a few of these cities the network did lose money and other city money had to cover the gap. But the answer is in the same section, where Baptiste writes that "the failures cluster around structure, not around public ownership, and structure is the variable a council actually controls" (Baptiste 274). What Baptiste means is that the problem was in how those systems were set up and not in the fact that a city owned them, and how a city sets something up is something a city can change. Thorsdottir makes a similar point when she says that "the risk of building is legible and the risk of not building is not, which is why the second one is usually chosen" (Thorsdottir 128). Meaning the cost of doing nothing is a real cost, people just do not see it on a ballot.

Baptiste also looked at what subscribers actually got out of it. In the survey portion Baptiste reports that "median resolution time for a reported outage was under twenty-four hours in the municipal systems and just over three days in the comparison group, a gap the subscribers themselves ranked above price when asked what they valued" (Baptiste 272). This shows the city systems fixed problems faster, and that customers cared about that even more than they cared about the bill. Baptiste adds that "accountability in these systems runs through a meeting that residents may attend" (Baptiste 273). What that means is if the service is bad you can show up at a public meeting about it, which is not an option you have otherwise.