I've been trying to fathom what exactly Neelie Kroes, Vice-President of the European Commission with responsibility for the Digital Agenda, intends to do about net neutrality in Europe for a while. Back in January of this year, I asked: "Will Neelie Kroes Defend or Destroy EU Net Neutrality?", and in June I was still wondering: "What's the Net Net on Neelie Kroes's EU Net Neutrality?"
Well, it looks like we finally know, thanks to blog post she wrote last week, entitled "Safeguarding the open internet for all." It begins well:
Huge innovation can be found online: we must safeguard it for everyone. Currently too many Europeans find that services are blocked and throttled by their internet provider. I believe every European should have access to the full and open internet, absolutely guaranteed, without such discrimination. And so I am coming forward with new rules to bring such practices to an end once and for all.
And I know consumers are fed up with being treated badly. Fed up with their operator retrospectively and untransparently changing their terms and conditions. And fed up with contracts that don't tell you what speed and quality you'll actually get. In a competitive, transparent market, consumers should know what they're getting and should be able to vote with their feet.
This rightly identifies that net neutrality is all about allowing innovation. But thereafter, things go quickly off the rails:
But, if you aim to protect an open network, overregulation is exactly the wrong way to go. The fact is, many innovative new services depend on fast connections over IP networks. If you want to invest in (say) new videoconferencing equipment, an IP TV, or a new cloud computing contract, you will also want to know your connection will support it. If EU laws banned such quality guarantees, we would risk effectively outlawing many of those new services too.
You don't need "over-regulation", you just need one extremely simple requirement: that every IP packet is treated like every other IP packet – no discrimination, in other words. "Quality guarantees" is simply marketing-speak for packet discrimination, which by definition destroys net neutrality.
But obviously, I can't just dismiss Kroes's claim that people won't invest in videoconferencing or IP TV, although I would note that both of these were around in the 1990s, which shows how backward-looking those examples are. But first let's look at her argument in favour of destroying net neutrality:
I agree that the Internet is an important platform for freedom of speech; and that is why I intend to guarantee access without restriction – the first ever time such a guarantee exists across Europe. But permitting premium services does not in any way restrict that freedom. For almost any kind of product you care to name – from postal services to the petrol in your car – having premium products hardly diminishes freedom for users: if anything it offers them more choices. If you decide not to pay that extra premium, of course, you still deserve a good product: and under my proposals, the "best-efforts" Internet will get better.
Again, it's rather significant the examples she uses here. Let's just assume for the moment that postal services and petrol in cars do indeed benefit from premium services. And let's now compare them to online services.
In the first case, you are dealing with industries that require enormous investments just to be a player: you can't deliver post or petrol without setting up national infrastructure. That means that even before you compete, you must spend large sums of money.
In the second case, Internet startups famously require little more than a couple of people in a bedroom with a PC and some free software. There are no other barriers to entry, no requirements for massive investments upfront, unlike in the analogue world of post and petrol.
Now let's consider the impact of premium services in both worlds. In the first, the cost of competing with a premium service is only marginally greater than competing with the rest of the market: the entrance costs are so high that increasing them to provide a premium offering is not significant. In the online world, though, competing with existing premium services require an investment that is likely to be much greater than the near-zero startup costs for non-premium.
Put another way, in the analogue world, premium services do not represent much of a barrier to entry for any company able to compete in the market at all. In the world of the Internet, premium costs are likely to add such a proportionately large extra cost that startups simply cannot compete without obtaining serious financing very early on. But it is precisely because Internet startups don't need to get that funding early that so many have been launched speculatively, and that so many have succeeded and then gone on to raise funds off the back of that success. In other words, the dynamics of the two worlds are very different.
This is the key difference that Mrs Kroes does not seem to appreciate. By allowing premium delivery, she permits incumbents or entrants with deep pockets to create barriers for startups that currently do not exist. She is actually negating one of the key strengths of the Internet: that innovative ideas can be tried out on equal terms with existing ones by practically anyone, without needing to ask permission or raise significant funds.
To make that more concrete, let's imagine that there is a company called Gopher Inc that allows people to access information on the Internet. Now suppose that there is some small startup called Web Ltd that has a great idea for making it much easier to access information online. At the moment it's only a system on a PC, but people who see it are impressed. As word gets out about this potential challenger, the well-funded Gopher Inc enters into deals with all the major ISPs and telecom companies to obtain premium delivery; that is, it is guaranteed that its traffic will always be delivered first and fastest.
When Web Ltd launches its hot new product on the market, early adopters like it for its revolutionary approach and cool ideas, but ordinary users find it runs so slowly compared to the tried-and-tested Gopher system, that it's just unusable. So they stick with Gopher, Web Ltd runs out of money and goes bankrupt. Needless to say, Gopher Inc buys up all its assets at a knock-down price to make sure no one else does, and then locks them up in a drawer for ever.
Notice that Web Ltd's service was neither blocked nor throttled – things which Mrs Kroes rightly wishes to forbid. But the fact that Gopher Inc was able to buy a premium delivery that made Web Ltd's service look dog-slow meant that in effect the latter was being throttled in comparison. This is a fundamental flaw in her approach: by allowing rich companies to buy better delivery, they relegate all others to the slow lane through a kind of Internet relativity.
Fortunately, when Tim-Berners-Lee invented the Web, he didn't need to worry about Gopher Inc. That was partly because Gopher never took off (but did exist), but mostly because net neutrality was simply taken for granted as a given. But once the new EU rules are in place, every open source, open content and open standards project would risk being sidelined by rich corporations offering closed products. Imagine, for example, if Microsoft paid every ISP to deliver IP packets from Windows systems faster than those from GNU/Linux machines: few companies would give free software systems another look.
Mrs Kroes goes on to say:
The fact is the Internet does not come for free. It relies on networks, and investment; if that does not come from taxpayers then it must be from private companies. If you prevent operators from providing what their customers might actually want — by effectively outlawing new services — you are making that job a lot harder. And that would be the wrong direction to head in.
It's nonsense to suggest that net neutrality "outlaws" such services. Anybody can create products that require fast internet connections – nobody is suggesting that they shouldn't, or that consumers shouldn't be able to buy such faster connections. But the key proviso is that all IP packets must enjoy that super-speed, not just those of one company – otherwise the Gopher vs Web scenario is a constant threat to innovation.
The issue is more whether the best network connections are fast enough for new applications. However, that has always been the case. When I started using the Internet, my first modem ran at 2,400 bits per second, which was a typical speed using the telephone network. When Netscape Navigator came out, it was optimised for the then increasingly-common 14.4Kbits/s modems. Today, most people in Europe can take megabit speeds for granted.
The point is, like processor speeds and hard disc capacities, broadband speeds are increasing by huge factors every year. There's even an analogue to Moore's Law here: the 1998 Nielsen's Law claims that a high-end user's connection speed grows by 50% per year. I'm not sure that's been observed as closely as Moore's Law, but certainly there is no reason to think that Internet connectivity speeds will hit a wall any time soon. The same is true for the backbone technologies, which continue to advance rapidly.
In other words, the necessary increases in speeds are almost certain to come through simply as part of the natural development of the industry. There is simply no need for net neutrality to be sacrificed in an attempt to make up for supposed "deficiencies". Once the high-speed applications start to come through, users will be willing to pay for higher-speed connections, which will thus finance the next network upgrades. That's how it works in every other industry, and its outrageous that the telecoms companies are claiming they need special incentives in the form of killing net neutrality in order to give them an incentive to do so. If they don't want to invest in their own future, they should get out of the business.
Mrs Kroes concludes:
Time and again I meet entrepreneurs who can see ahead to tomorrow's online opportunities – from secure cloud services for business to eHealth and telecare for an ageing population – and they agree on the kind of networks and quality guarantees those innovations need. Helping them to grow is not about "helping telecoms companies ": it's about helping everyone who could enjoy and benefit from those new services. And I do not intend to be the one who kills off those opportunities.
But that is precisely what her proposals will do. By allowing premium pricing she not only destroys net neutrality, but she also puts back the barriers to entry that the Internet's rise has swept away from countless industries. She ensures that only companies with deep pockets will be able to launch new services that challenge existing ones. And she places all open projects – the real engine of innovation today – at at a huge disadvantage, since they lack the funds to buy premium delivery that would provide them with a level playing field.
Most ironically of all, she ensures that if the next World Wide Web is invented in Europe by some small startup or lone genius, as the current one was, it will never succeed: her death-blow for European net neutrality gives sclerotic but rich incumbents the right to pay ISPs and telecoms to have dangerous innovation consigned to the Internet slow lane - and thus oblivion.