The argument for forced sharing of the last mile isn’t that it’s a natural monopoly – that’s unempirical but commonly claimed. (Prior to FedEx and UPS, smart people thought package delivery was a natural monopoly.)
The better argument is that the providers of last mile are beneficiaries of prior regulation which meant that they were able to build a gov’t-protected monopoly.
The latter is a legitimate but, to me, not dispositive. We want competition on the last mile (and every mile) and line-sharing regulation simply cements incumbents, as no new entrant would desire to be in a business of regulated rates.
Now, we might simply accept that prior damage has ruined that market beyond repair and the least-bad solution is to force line-sharing. But I’d rather see lots of Sonic’s building a new set of last miles.
Suburbia in general is only possible with supportive regulation. There is a reason Sonic's business model has only developed in very dense neighborhoods (4000 person/mi^2 vs <1000p/mi^2 for other suburbs).
If we are going to live in economically infeasible locations, relying on government subsidy to pay $30m/mi to build highways to shuttle us to work, then we are going to have to rely on subsidy for other aspects of our lives as well. The regulation/subsidy-free ship sailed about the time we started the national highway project, collectively bought an automobile company, etc.
There's little to no evidence that the Interstate Highway System couldn't be self-sustaining if that were desirable. As it is, the federal government decides instead that about 1/6 of the money raised by fuel taxes and other fees should go to public transit, and states prefer to get their funding contribution from property and other taxes.
Well, of course Naderites are going to have that perspective. But stuff like this isn't particularly persuasive:
> Since 1947, the amount of money spent on highways, roads and streets has exceeded the amount raised through gasoline taxes and other so-called “user fees” by $600 billion (2005 dollars), representing a massive transfer of general government funds to highways.
$600 billion 2005 dollars over 64 years? That's an average of less than $10B/year for a system that serves literally every American, either directly or indirectly through the goods that are delivered by trucking.
Even if road spending has been growing faster than the rate of inflation, that would still be a bargain at twice the price. If I were a properly motivated advocacy group, I bet I could produce a number with a T at the end of it and somewhat plausibly claim its the economic activity currently dependent on the asphalt transportation system.
It takes a real ideological bent to call that a "massive transfer of general government funds". After all, some of those general government funds are filling in gaps left by drivers subsidizing train, bus, and subway passengers. Presumably a not insignificant chunk of the rest represents municipal costs, where it's hard to use tolls or gas taxes to build local roads that literally everyone uses even if they're just biking to the train station.
The cost per user to support 1,000 people/mi^2 vs. 4,000 people/mi^2 is not that dramatic. At first glance the distance between houses might seem to be the square root of the population density, but people tend to group together. Picture 2 sets of row houses one of which has a back yard and the other does not. As the ISP's run their wires down roads it makes vary little difference how big those back yards are just how far down the road you need to travel from one house to the next, and how far from the road to each house.
PS: If you’re on city water chances are you’re fairly cheap to wire up to a modern network. Unfortunately the incumbent players tend to also be selling you service on the old network so they have vary little incentive to upgrade you especially if it might lead to you dropping some extras.
It's been well established in the academic literature that infrastructure costs in America cities are related by a power law to population. Costs increase about 80% with a doubling of the size of a city. So we generally would expect infrastructure in a city four times the size of a typical 1000 people/mi^2 suburb to cost about 3.24 times as much, or about 81% as much on a per-capita basis.
There are many well known economies of scale to denser urban agglomerations that suburbanization forfeits. That's not controversial. What is controversial is why suburban living, despite its much higher cost compared to city living, has nevertheless come to dominate post-war America.
In most countries, the rights to lay those cables are highly restricted by law, and while that reality is perhaps not ideal, it is also unlikely to change.
So I'm not arguing that it's a natural monopoly, just pointing out that with the existing government regulation that limits the number of entities that may run last-mile cable, we won't see effective competition unless those restrictions are balanced by requirements for shared access.
I too would love to somehow have a whole bunch of companies able to lay cable right to my house, but I doubt that will happen. Managing such right-of-way is actually a very hard problem to solve (which is why governments regulate it in virtually (literally?) every country. So I think that's an unlikely outcome.
The better argument is that the providers of last mile are beneficiaries of prior regulation which meant that they were able to build a gov’t-protected monopoly.
The latter is a legitimate but, to me, not dispositive. We want competition on the last mile (and every mile) and line-sharing regulation simply cements incumbents, as no new entrant would desire to be in a business of regulated rates.
Now, we might simply accept that prior damage has ruined that market beyond repair and the least-bad solution is to force line-sharing. But I’d rather see lots of Sonic’s building a new set of last miles.