A thousand cuts and the last mile problem

Tim B. Lee has been writing an excellent series of posts on Net Neutrality at Vox. See here, here and here. The picture painted is of the structure of the previously decentralised internet in the US being a mess and one where those with some degree of power (in this case afforded by size) are trying to carve out their slice of coherence raising the spectre of future market power, if not market power today.

Something I will write more about soon is that I must admit that as you read more about this, you become pessimistic that regulations — such as Net Neutrality or a variant — can be effective against the myriad of options to get around those rules. More and more you get push towards the idea that getting the structure of the internet and competition right would be the way to go. After all, Net Neutrality is hardly an issue elsewhere even though there are surely similar drivers with regard a tension between the use of price discrimination and the different ways of funding infrastructure.

Today’s piece in Vox, however, reminded me of something important in this regard. At the heart of the problem with competition on the Internet and how this might impact on consumer choice is the ‘last mile problem.’ The issue is that there is a single ‘pipe’ running into dwellings and premises. To have more than one is not cost effective. And that ‘pipe’ will always be owned by someone and that someone will have monopoly access to the consumer as a result.

This is not an uncommon problem. For example, it was a problem for telephony well before the Internet. Initially, in the US and elsewhere the last mile monopoly problem was solved by having no choice at all and having an entire network monopoly problem. But following the break-up of AT&T, there were multiple networks in the US. That meant that for long distance service, those monopolies now mattered. However, as we expected call traffic to be balanced (there is some simple maths that shows this), those networks who were, before their forced breakup part of a single culture, worked out pretty quickly that, at least between them, agreeing not to charge one another — so-called bill and keep — was the way to go. After all, it would all balance out. As the Internet became grafted on to the existing infrastructure, that same set of arrangements proved convenient.

In other parts of the world, previous monopolies were not broken up but the networks were opened up. None of the incumbents wanted bill and keep and so they resisted and then were regulated. Basically, they were forced to provide access to their competitors to the last mile at regulated prices. This proved to be (a) quite attractive to new entry and (b) not a bad deal for incumbents as they converted the last mile into a regulated priced annuity. The end result was that there was enough competition for consumers that there was no one with enough size or incentive to both with trying to overturn Net Neutrality. Instead, there were some limits on total use which has its own issues in terms of market power but at least special deals with content providers were largely off the table.

So what are the options for the US to deal with the last mile?

1. Regulated prices: it could adopt regulated pricing as in other countries and open up access.

2. Privatisation: the essential problem with the last mile is that there will always be a last mile because there is always a customer at the edge of a network. The problem at the moment is that the customer is a bottleneck and then the owner of the last mile to the customer is a second, sequential bottleneck. Essentially, the double monopoly problem is built into this structure. What if, instead, the customer owned the last mile or at the very least the cable into their home? That may allow for easy switching of their providers without having to duplicate or appropriate property. And customers are surely the natural owners here anyway.

3. Municipal broadband: saving the customer owning the last mile, individual localities could set up their own networks to provide a second option against incumbent firms. This is the ‘thousand cuts approach.’ Basically, what we have in broadband is a number of monopolies equal to the number of customers (dwellings/premises). The fight is over each one individually. One approach would be for a large national competitor to emerge and do this everywhere. But what would make more economic sense is that there were local solutions based on local needs. The type of solution for a city is different from suburbs and is different again from regional areas. Some may require wired solutions whereas for others wireless could work. This last scheme of getting competition at the local level on broadband was something I pushed for in Australia. (They went for the government funded nation-wide solution that seems to me not to solve the underlying issues and cost a lot but that is another story). But the idea was proposed by Derek Slater and Tim Wu for the US.

There may be other options as well. In any case, I think it is time for the US to start thinking beyond letting regulation in the form of Net Neutrality do all the work for what is fundamentally a structural issue.