Even before passage of the 1996 Telecommunications Act, “non-dominant” access providers had been using their rights to “cherry pick” where they build communications facilities for some time.
First “alternate access” providers got into business to provide high-bandwidth connections for business district enterprises and their designated long distance providers.
After passage of the Telecommunications Act, competitive local exchange carriers used a combination of leased and owned facilities to provide voice and data services to business customers.
Then Google Fiber convinced city regulators that allowing Google Fiber to build gigabit access facilities neighborhood by neighborhood, and only where there was demand.
That might well be one reason why Verizon suddenly reversed course and decided to build FiOS in at least some neighborhoods in Boston, where it previously had held off from doing so.
Basically, “cherry picking” (selecting to serve only some customer locations) became lawful in an ever-widening number of cases. That new flexibility comes on top of the typical “ubiquitous build” required for the incumbent telcos and cable TV companies.
Eventually, harder questions will have to asked about how the social function of “universal service” is satisfied, when there is no dominant provider with monopoly profits to “tax” for such purposes. We already have taken a half step away from the notion. ISPs and competitive local exchange carriers other than the former incumbent telephone operator have no universal service obligations.
But what happens when the former incumbents are clearly not dominant, because they have lost so much market share?
Assuming some form of universal service requirement will be imposed, in some way, on service providers, the issue is how to apportion the cost among a range of providers, none of which truly is “dominant.”
It is not hard to imagine markets where three fixed networks and three mobile networks all provide high speed access, and eventually all six networks can offer gigabit service over at least significant portions of their coverage areas, and none is “dominant.”
Collecting the tax is not difficult. We already tax telecom bills. We already disburse funds, in rural areas. In urban areas, there has not traditionally been an issue. Eventually, there will be.