If CenturyLink is allowed to buy Level 3, then California will lose a major source of dark fiber. That’s my reading of a settlement agreement between CenturyLink, Level 3 and three of the organisations that challenged the deal at the California Public Utilities Commission. The fourth challenger, the California Emerging Technology Fund, isn’t part of the agreement.
What I don't get is why would someone not want to lease this unused capacity? Is it to squash competition and push up consumer costs?
Backbone data prices have continued to sink, year after year. "Last mile" connectivity prices continue to rise and far outpace the underlying data transmission costs (and particularly, the backbone portion of those costs).
If you have spare backbone capacity, but are also in the very lucrative business of providing end-point / last mile connectivity, why would you lease your excess capacity and open yourself up to competition in the lucrative last mile segment?
P.S. Take "last mile" figuratively, in the above. I don't know where the line exactly between "local"/lucrative and "long haul"/ever-cheaper.
Enterprises can outsource management of the capital equipment (wavelengths) or the entire network (e.g. Ethernet). The latter is known is lit services and is much more common than dark fiber.
Overbuilding is not the differentiator though. During the original dot com boom of the 90's yes there was a lot of excess capacity due to overbuilding but dark fiber is really any fiber that is currently unlit that is available for lease.
Building capacity can be and often is the business model itself. For example you pull fiber through an urban area and then sell IRUs(indefeasible right of use) on a cable pair, usually for a period of 20 years. As a fiber provider I may not have ever intended to light that fiber myself so it was never overbuilt.
The largest cost of laying fiber is not in the fiber itself, the glass itself is relatively cheap in comparison to the cost of obtaining the right of way, the labor and so on. Often municipalities, counties, states or whatever charge a fee for these rights of way if the fiber is underground (the most expensive). If they are attaching aerial fiber to existing poles, the pole owners, often power companies, have also paid for the rights of way and further charge a poll attachment fee. Regardless, all the other cost typically far over shadow the cost of the individual strands.
Typically a carrier will install a larger fiber count cable since the incremental cost of installing a 144 count fiber rather than a 72 count for example, would probably be relatively minor so why not have a few extras. Most carriers have the choice as to whether or not to sell off (typically using an instrument called an Indefeasible Right to Use, which is sort of like a condo but for fiber, usually long term, 20years). Usually they will sell or trade with other carriers to help offset costs, for example carrier A might trade 4 fibers in San Francisco for 4 fibers in LA for particular routes. In that way they are not so directly aiding a competitor but also gaining routes for themselves.
Usually, carriers do not like to sell off DF to customers such as large enterprises because quite simply, those customers would never have to buy anything from them again because they could continuously gain more capacity by placing DWDM equipment on each end (not endlessly of course but for all practical purposes for most customers, endless).
That is why most carriers like Century Link or others (especially ILEC's, Incumbent Local Exchange Carriers, i.e. the phone company) do not like selling DF. Typically these carriers have purchased a Franchise agreement with cities or whichever authority. These franchise agreements are not cheap (usually 5% of gross or something like that, basically it is like a tax) that gives them in theory access to all rights of way from the controlling body (cable co's do the same thing).
Because of that they maintain often a de-facto monopoly or a powerful market position that they do not want jeopardized by the availability of DF which would provide extremely cheap bandwidth.
I have often thought that the smart thing for cities to do, if perhaps one were to be able to start from a green field, would be for the city to place telecom conduit up and down every street from the get go and then rent out inner-duct space to carriers. That would make it easier for competitors to access the rights of way without all of the disruptions.
But perhaps new mmwave 5G technology will eliminate such concerns eventually (but I am skeptical).
