Anatomy of a train wreck: 5G and net neutrality
Following on from my article suggesting net neutrality in the US might become a slow motion train wreck, I wondered whether Europe was going to have a train wreck too. I need to examine more closely the arguments of net neutrality proponents to defend why I use the train wreck label to describe debates around net neutrality and 5G.
I've written a number of times why I think paid prioritisation could be key for getting 5G deployed. Lots of people will continue to comment on that and I don't wish to further consider it here. Instead, I write about the other side of the coin: why is paid prioritisation so difficult to accept for proponents of net neutrality?
The phrase network neutrality is widely attributed to lawyer Tim Wu. In a famous 2003 paper, Wu outlines evidence that broadband networks in the US commonly engage in discrimination of the traffic they carry. Discrimination is important in order to maintain networks and manage bandwidth. Yet, he argues that it detracts from Darwinian competition in which only the best survive.
Wu's is the seminal article on the issue and he admits in it much work remains to be done, While he identifies issues like restrictions on inter-network traffic, he fails to offer a convincing definition of the big concern. For example, did we really start out living in some Darwinian wonderland of innovation? Many before have written about the path dependency in technological innovation: successful systems like China's WeChat or Kenya's M-Pesa haven't developed in the West, Moreover, Wu admits lots of discrimination occurs for benign reasons of network management. How can we distinguish the good discrimination from the bad?
In spite of the issues of definition, Tim Wu identified a problem that resonated strongly with many people: some proponents feel very passionate about it. When Berec consulted on network neutrality in 2016, it received almost half a million responses, an unprecedented number. For many proponents, moreover, only blanket bans will do to cover all behaviour they regard as contravening net neutrality. For example, this 2015 article by journalist Jon Brodkin quotes those who worry about loopholes in European law on the issue: with those in place, what might telcos manage to get away with?
I believe this shows what makes paid prioritisation so difficult to accept for proponents of net neutrality. If it's allowed, what else might telcos get away with? If telcos benefit from paid priorisation, won't that give them an incentive to engage in throttling and blocking?
I can now justify what my train wreck description stems from: the passion of net neutrality proponents on the one side and, on the other, the possible need for paid prioritisation to fund telecommunications services. A train wreck is a possible outcome because a wedge gets driven between these two sides by the difficulties defining the problem. Some commentators, like Tim Lee, get frustrated with the difficulty of pinning down what proponents are most worried about. What I think it shows is a need for learning on both sides of the debate about the environment that develops, if and when paid prioritisation starts to get introduced.
It seems to me that learning about the developing environment is the only thing that can remove the wedge between the two sides of this debate. In that sense, I argue that the most positive net neutrality legislative changes that might come about will be those that promote openness from telcos and content providers about the agreements they enter into, while not unduly restricting what those agreements are about. We should readily believe we can rely on standard competition law restrictions on firm behaviour to control inappropriate behaviour in the mean time.
Indeed, a standard competition policy approach to paid prioritisation does have its advantages. For example, I haven't mentioned yet the difficulty of pricing services in two-sided markets of the sort paid prioritisation might create. Might a margin squeeze test provide an appropriate way for telcos to price paid prioritisation, when they have their provide content services, as well as carrying 3rd parties'?
Evidence about this issue will continue to develop. If you feel like following what's getting discussed, including by me, please come and join my group, Commercializing 5G.