I went to Wednesday’s Westminster eForum symposium with a single, simple agenda: gather ideas on how telcos play politics. Top-dog regulator, top-dog political master, top-dog incumbent. I got what I paid for, and I’ll be incorporating some of what I learnt into the advice I give my consulting clients.
But the debate also disappointed me. I don’t think it got to the most important issues, and nobody asked the really hard questions. This may be because the debate is predicated on false assumptions.
So let’s look at the assumptions behind the process that led to the OFCOM, DTI and BT pitches. The starting point was the Phase 1 strategic market review by OFCOM. This concluded that there remain market failures in UK telecom that require fixing through improved competition.
Reframing the debate
But is market competition even in principle capable of delivering a satisfactory outcome? What is an acceptable outcome for a service that, as Ben Verwaayen notes, forms an input to almost every other economic activity? And even if competition can deliver, as a matter of public policy should private entities have such control over the public’s conduit to information?
As a hardened free-marketeer, even uttering the question makes me shudder. But the libertarian inside me also quakes at my free speech depending on a single commercial supplier. And the pragmatist tells me the dribble of connectivity into my home isn’t keeping my small business competitive either.
Telecom is not a normal market good. Let me elaborate.
Connectivity isn’t a can of baked beans
Traditionally economists have identified four core factors of production that come together to create markets: land, labour, capital and enterprise (“entrepreneurial drive”). None of these is run as a standard market based only on competition, contract and tort law. They are restricted and shaped in the public interest by rules such as planning permissions and zoning, employment law, financial services regulation, and company and tax law.
I would contend that connectivity is an equal of these four, as information goods come to contribute a dominant share to the value of products and services. And just as the special characteristics of land and labour have resulted in special rules, it should come as no surprise if we do the same for connectivity.
How is telecom special? It’s the one industry where less can be more, since stupider is better. In most industries adding features and services increases consumer value. In telecom it frequently erodes value because they’re just price discrimination mechanisms. Vertical integration locks out rival innovations and competition to the detriment of the economy as a whole. Furthermore, the very nature of network effects makes these entry barriers very high.
Click your heels, dream of a better place
“Stupid network” connectivity services can be defined by a relatively small number of non-functional factors (reach, speed, latency, etc.). Their implementation may be subject to rapid innovation, but the factors themselves are well-understood, complete, and invariant. We could have a meaningful debate about the type of connectivity we’d like to have in 50 years from now; yet we couldn’t anticipate the most important functions of a computer processor 25 years from now. We might not have even invented terms to describe those functions yet.
Telecom isn’t alone. Take water. We know what clean water looks like, and we know where we want it delivered. There’s no mystery in that.
Normal markets are a bit like Richard Dawkins’ blind watchmaker model of evolution. We don’t have an end in mind because nobody know what ends are possible. We just make locally optimal incremental steps, and hope for the best. Central planning fails because no one mind can hope to encompass all the market information and imagination to make optimal investment decisions.
Connectivity is different from cornflakes. We can have a debate about outcomes, since we don’t need a market to explore what the possible outcome space might me. On an exceptional basis, you can have a public policy that specifically aims towards a target state. This is precisely what the Koreans, Malaysians, etc. have done. It’s like Eisenhower and the interstate highways: you’ve just got to get it done to compete, even if you have to hide it in the military budget.
That means the OFCOM director is wrong and the polticians were right: competition is a means to an end in communications, not the end in itself. Being technology-agnostic is not the same as being connectivity-agnostic. Competition is a means of selecting between technology implementations, not connectivity destinations.
If you want to make the man from OFCOM uncomfortable, ask him if he’ll stake his salary on Britain having the best connectivity in world in 10 years from now, and watch him squirm. Would he claim that being anything less than #1 is a triumph? How far down the list does he deem appropriate?
BT isn’t an evil monster, just a dangerous zombie
In hindsight, the BT privatisation back in the early 1980s was a screw-up. The ownership of the local loop was transferred from the public to a private organisation. At the time the value of that loop only reflected the potential of phone and fax services. Later along came the idea of broadband and a stupid, flexible, pervasive Internet. Initially this had to be resisted by BT, as the services could not be metered and it disintermediated the voice product. Over time there has been an explosion of service possibilities. The value of loop has rocketed to reflect that, but not as a result of any innovation from BT. There is no claw-back clause to reflect that and return it to the public. It’s a pure access rent. Even without supernormal profits from uncompetitive markets it starts to look attractive to deploy dumb pipe connectivity. Better to capture 25% of a $100-valued service than 100% of a $20-valued service.
So BT has had a massive unearned windfall. But it can only capture the the full value of that pipe by metering and charging for the traffic that passes through it. BT’s purpose in life is price discrimination of traffic passing over a largely fully amortised network. (The 21st Century Network is just a way to create a closed, tolled private network on cheaper IP hardware with low opex costs. It isn’t about creating a stupid network for users to take control of.)
Nobody asked the BT CEO how he plans to make money from his new network; what will it do that a vanilla Internet connection can’t do; how will he stop people bypassing his service toll booths?
What’s the price of free speech?
Freedom to connect is not a market good in the same way we don’t charge people to enjoy freedom of speech or freedom of movement.
Would we tolerate for a moment a British Roads Corporation that inquired as to the destination and purpose of each journey and charged accordingly? No! We’ll accept localised congestion charging on a non-discriminatory basis, but not general system funding via individual price discrimination.
That doesn’t mean printing presses, cars or computers should be free. But the rules of the game may exclude certain rent-seeking behaviours and privacy invasions aimed at price discrimination.
Can the market deliver?
It’s quite possible that BT will never deliver the desired type of connectivity, by the very nature of its constitution. They must always sail as close to the regulatory rocks as possible; get too far away and the board kick out the CEO for one with a more acute sense of navigation. The regulator can’t push things too far, because they’ll be out of a job. There’s mutual co-dependence, which as the ntl VP noticed works to re-inforce existing market structures.
The market delivers roads and railways. But it can’t decide where those roads and railways get laid. You just solicit bids for implementation and operational management. We need to re-think what the market and regulation are for. The model is broken, not the implementation.
What to do about BT?
How to get to “there” from “here”? Direct buy-back of the local loop is too expensive. Even if BT doesn’t deserve the rent, that was the deal with the shareholders. You can’t just take it away; in the interests of preserving property rights and an orderly society, you don’t want to go there. So we instead have to resort to hidden expropriation via regulation. After all, this is a risk the shareholders knowingly undertook.
More access networks, please
One useful idea to come out of the proceedings was that regulated prices should be set at the indifference level of a new entrant building an access network. They should not be based on the bloated costs of an incumbent. However, I would go further. I do not believe we should be agnostic between platform competition and open access. Platform competition is superior because it doesn’t rely on political caprice. It’s hard to reverse. It naturally leans towards more competition and less politics. It stimulates new technology. So I’d force the price paid to access unbundled elements to be a bit more than the cost of building your own. The excess wouldn’t go to the incumbent, however; it would go into the general fund to promote universal service, technology research, and cover the costs of what regulation does occur.
I’d also make it easier to create user-owned access networks through spectrum liberalisation and open access to conduits. BT’s wayleaves across the streets are not private property of BT; they’re granted by the public and the public is entitled to appropriate recompense.
Outlawing price discrimination by the back door
The playing field needs to be tipped to create disincentives to price discrimination. Think of the 3G auctions; sure the spectrum is worth $40bn to the users, but that doesn’t mean the operator will be able to appropriate that value. If “stupid” connectivity had been mandated from the outset, the greedy operators might not have been raped so badly. Their spectrum fees, if any, would only have reflected the confidence they had in the superiority of their technology and implementation plans over the next-best rival. It wouldn’t have represented their ability to tax all wireless speech.
Internet Protocol is blessedly content-neutral, and it’s only when you get to the TCP and UDP protocols you get service-specific port numbers. Want to host a web server? That’ll be extra Sir — we don’t allow inbound TCP segments with the SYN flag set…
There’s a simple way round this. Amend the wiretap laws. Anything above IP is verboten. Your carrier should carry, not look. Call it the Medusa principle — directly looking at the data will kill you. If the most you’re allowed to see is a source and destination IP address, it gets damned hard to price discriminate. Oh, and you wouldn’t be allowed to discriminate based on IP address range either.
OFCOM’s false trichotomy
The OFCOM choice of deregulation, break-up or entente cordiale is a false one. Asking BT what to do is ridiculous. Their job is to defend their windfall. Ofcom’s job is to return it to the public. BT’s rightful goal is to be one iota away from forced breakup, and not give one unnecessary inch of concession. They’ll get “option 3”, and business as usual. Where’s option 4 – realignment of ownership and incentives? Not structural separation, but structural re-design? Breaking up BT without addressing the wider policy issues on connectivity is pointless and destructive.
The failure of the debate is to ask what the appropriate ownership and management structures for public networks should be. Increasingly, user-built networks like Wi-Fi and community meshes are being deployed. Networks deployed by users, communities or municipalities better align the interests of owners and users. There’s less need to price discriminate in order to generate supernormal profits. The benefits to the wider economy should greatly outweigh the losses to the telecom market.
The fire escapes are locked
BT probably has more lawyers than OFCOM has staff. Nobody’s going to tackle them head-on; BT can threaten job losses and service troubles, and parliament will do its bidding accordingly. The political will isn’t there to re-examine our industrial policy and make the UK the hub of the global information society. No regulator has ever declared the job done and quit.
The only way out for now is a war of attrition on the technocratic rules of engagement, one by one. Eventually a slowly rising tide of stupid connectivity erodes BT’s service revenue base. The connectivity cartel will be broken, but at what cost?
At the very least we can try to get a more sceptical audience for the next public engagement and give the players a harder time. Any volunteers?
Posted by Martin Geddes at 10:57 PMTrackBack URL for this entry:
http://www.telepocalypse.net/cgi-sys/cgiwrap/mgeddes/MT/mt-tb.cgi/386.
Thanks for that Martin, very neatly expressed and even a little "freedom" as desirable in the mix.
I very much agree that selling the last mile (and the numbering plan apparently here in NZ) was the fundamental flaw in selling the incumbent.
And good to see someone raising, amidst the flood of property talk, the degree to which incumbents receive corporate welfare (sponsorships of new infrastructure in many jurisdictions) and have rights that others don't share, while they exploit the populace they would have you believe they serve.
As for regulators/incumbents, it all ends up with the Stockholm Syndrome, they are, lets be honest, co-dependent.
Yes, wireless is going to hammer the last mile, while the incumbents, particularly those who could drag their bulk to the cellular trough, are doing their "triple plays" and other vendor life-preserver sales.
The other step is to move away from the exchange, that's the true value of the last mile, it leads to a market of a single supplier. Lets take our wireless clouds and swarms to open neutral peering exchanges and choose our own transit providers and peer promiscuously.
Of course the Intellectual Property Government complex is going to oppose that freedom, for their own reasons but for the shared goal of power and control, but we can win.
Just keep blogging, flickring and soon the centre cannot hold.
You describe the privatisation as structurally wrong. I take it you would have preferred the local loop to have been kept as a separate (not for profit?) common carrier. Where it also went wrong is in what happened to the money raised by selling off BT. In practice it was used, as any economist will tell you, to keep the level of tax lower for a while than it would otherwise have been. What it SHOULD have been used for is investment in the local loop. This would have enabled the use of a nationwide, fibre to the curb (at least-possibly end to end fibre) network as envisaged in an extraordinarily visionary report from NEDO in 1987 called IT Futures: IT can work. !987 being pre internet of course.
They had the vision to recommend that a nationwide fibre optic grid be installed. The Government didnt agree. Apart from it being profundly against the spirit of the (Thatcherite) times, no one could think, it was said, of any applications that would make use of such bandwidth. We are talking Gigabits of course, with the technology available then...
Posted by: at October 17, 2006 04:49 PM