Why network effects require ongoing interconnection rules (and other regulatory interventions)

I write today’s blog from sunny Switzerland. The news seems rather slow and I can’t bear to write a top ten lists of things that either happened in 2010 or to watch in 2011, so today’s post will try to explain what network effects (or externalities) are and why even in a minimalist government intervention capitalist society telecoms networks require ongoing interconnection regulation. Sounds a bit dry, but stick with it – the same principle can be applied in other circumstances and once you have the tool in your box, you’ll be surprised how often it can be used to help to explain otherwise strange regulatory interventions.

We’ll start with a thought experiment (no cats will be hurt in this one). Imagine a small town with 12 inhabitants. There are two telecoms networks – network A has ten customers and network B 2 customers. It is an unregulated town and although network B has asked network A to connect the networks together, network A has refused, so customers on each network can only call other customers on that network. You arrive in town in your zero-emission car and become inhabitant 13. Network A and B both offer you telephone service – identical price and conditions. Which one do you take service from?

Most people choose network A as they can call more people. As most newcomers join network A it keeps getting bigger and the pull becomes stronger. This is a network effect or externality, and is often described as ‘the value of joining a network is proportionate to the number of people connected to that network’.

Within telecoms, network effects are to some extent neutralised by the impositions of a rule requiring the two networks to connect their networks together so that the thirteenth person can call all 12 people in town regardless of which network they join (this also has strong social benefits – and this is sometimes described as the ‘any to any’ principle).

Of course, without more a bare interconnection obligation doesn’t address the unequal bargaining power of networks A and B, nor some of the other advantages network A might possess such as economies of scale, scope or network density, but I’ll come back to those another time.

This is well-understood in telecoms regulations, forming part of the WTO reference paper and forming part of EU and Us telecoms telecoms policy.

However, any network can also show network effects. The balance that regulators and governments need to address before intervening in those markets (particularly in emerging markets such as social networking) is whether they also have the same characteristics as telecoms markets of large capital requirements, sunk costs and incumbency that create barriers to entry in telecoms markets.

[polldaddy poll=4315674]