It costs you $1,000 just to pass a house, even if they don't subscribe, and you have to maintain that infrastructure too. If 40% of passed houses subscribe (Verizon has been struggling to hit that number for years), you pay $2,500 per house that does subscribe. This is the "uptake ratio" and a fiber provider lives and dies by that number.
Build-out requirements tank your uptake ratio, because you spend a bunch of money passing houses where people can't afford to subscribe. It increases how much you need to make on all the other houses to recoup your costs.
But your other subscribers don't give a shit about your costs. They're willing to pay, e.g., $100/month for triple play, and they don't care if your per-house cost is $2,000 or $4,000. If you raise your prices to recoup your costs, they'll just stick with the incumbent, again tanking your uptake ratio.
A small ISP may be willing to ask for a $10 million bank loan to wire up a yuppie neighborhood like Fells Point, with the prospect of seeing how it goes and building out from there. Nobody will ask for a $250 million bank loan to wire up the whole city of Baltimore.
A city that wants fiber has the obvious choice to subsidize the investment of a large ISP, say Comcast, thus severely reducing the investment cost. The permission to build is of course non-exclusive, but the subsidizing is not. Once built, the ISP gained a government granted advantage, and thus a monopoly is born.
Alternative, a city can create a government bid for administrating and building a city fiber network. On paper, this mean that the city owns the fiber, and the ISP that builds it has to sell access on equal ground with competitors. The ISP do still get an advantage in operation knowledge, some payment for doing the administration, and first mover advantage. This tend to result in healthy, if somewhat constrained, competition.
Ars Technica narrative talk about how the cable division paid little or no construction costs. That is very different aspect than if the permission to construct, operate, repair, maintain, and reconstruct a Cable System is non-exclusive.
Cities don't subsidize cable companies. Indeed, they extract concessions from them. E.g. every time Comcast re-ups the contract in Wilmington, DE, where I used to live, they have to kick in a random couple of million to government programs.
> Ars Technica narrative talk about how the cable division paid little or no construction costs.
Who cares what division within the ISP paid the construction costs? At the end of the day, the money came from the company's paying customers, not the municipalities.
A second mover really only has a chance of getting 50% of the market. But it still has to build a network that services 100%.
A third mover really only has a chance of getting 33%, but has to service 100%.
The first mover doesn't get subsidies, they just get 100% of the market. The second mover likely can't be profitable. The first mover will match or beat their prices.
Why is that pre-condition supposedly preventing companies from building infrastructure? It's not as though the City has also demanded that they operate at it a loss, merely that they build out service to the entire city as opposed to building service out to the choicest bits. And, isn't that exactly what a city government ought to do?