Broadband Access - What should we regulate?

Earlier today I spoke on the "Access Reverticalization" panel at VON Europe Spring 2006.  The subject was first mile access and the panel was effectively rigged, i.e. there was remarkable the level of agreement among the speakers.  My presentation is here.  Between the presentation and the Q&A, my points were:

1.  Regulation & legislation: be careful what you ask for.  Despite the best of intentions, it's easy to get unexpected results.  I used examples I've previously discussed here & here & here. 

2.  Laws and regulation evolve slowly - over years and decades - so it's worth thinking about where we want things to be in the long term.  Indeed, it's a bad idea to argue about evolving today's legacy telecoms regulation if we don't understand what makes sense for the long term, i.e. decades.

3.  Competition works, even in telecom.  This is clear if we examine relative mobile subscriber growth in the more than 100 countries that now have mobile phone service.  Three viable competitors yields growth, while four or more competitors generate really exciting growth.

But how should we choose what to regulate and where to encourage competition?

4.  Look at performance evolution rates at each "layer" of the local connection.

• Routers and Switches, i.e. TCP/IP or Ethernet: Faster than Moore's law
• Lighting the fiber:  Faster than Moore's law
• Fiber itself:  Slowly improving - 20 year useful life
• Conduits and poles:  Slow evolution - 20+ years useful life
• Local rights of way:  Fixed, limited and communally owned.

This table suggests an answer!

5.  We have a real bottleneck in the limited local right-of-way, so regulation is required.  But it makes sense to match time scales.  That means we should regulate at the stable layer and foster competition at every layer above, certainly at every layer above dark fiber.

How might this work?

1.  Competitive access to poles and conduits, a.k.a. layer zero competition.  The government controls the poles and conduits and makes them available to any interested party including developers, service providers and community or private organizations.  With overlashing on shared support cables, many more cables can fit overhead than is typical today and additional conduit banks can be deployed on popular underground routes.

2.  Condominium fiber.  Developers, or contractors working with groups of interested parties, lay fibers and sell individual fibers on a condominium basis.  There's a good overview of the subject here.

3.  Municipal dark fiber.  The government organizes the fiber deployment and makes dark fiber available to all comers.  Stokab AB in Stockholm is a very successful example of this model.

None of the above requires that individual homeowners worry about these issues.  The point is to make it easy for many, many providers to compete to offer services to individuals and businesses.

Is it possible?

There are many vested interests that stand in the way of realizing anything like this, but if we don't articulate the desired end goal, we will never get there.

Comments

[Razib Ahmed]

Since, you are interested about mobile phone and internet in South Asia, you will be happy to know that Pakistan government has taken an iniative to connect 1.5 million to 2 million households with high speed Internet in the urban areas. ALso, today, Pakistani Minister for Information Technology, Awais Ahmed Khan has stated that Pakistan has now 28 million mobile phone users. (http://www.southasiabiz.com/2006/05/can_pakistan_create_200000_job.html)

[brough]

Razib, Thanks for the reference. I read your blog regularly, but I've been days & days behind. Checking further on what Awais said, it appears he's got a plan to use their universal service fund (USF) to subsidize rural broadband deployments. I don't know how USF works in Pakistan, but in most countries it's a ponderously slow way to accomplish anything.

[Razib Ahmed]

Thanks for telling about USF. I do not know about it much and I am trying to get information about it more. Well, you are an expert on communication. The main problem I see about implementing broadband in Pakistan is that in the rural areas density of population is very thin compared to Bangladesh. So, for Pakistan wireless broadband will be better.