Monday, 7 March 2011

Why the Liberal Submission on Usage Based Billing is an Epic Fail

Lies, damned lies, and politicians....

I read Marc Garneau's submission to the CRTC this morning and nearly had an apoplexy.

This submission is nothing more than populist generating clap-trap that isn't worth serious consideration by anybody. So therefore it is an official position of the Liberal Party of Canada. Useless 'effin gas-bags the lot of them.

Anyway, so lets do a point by point deconstruction of the arguments presented:

Point 11: "We fundamentally disagree with the Commission’s description of the Internet as analogous to a utility or the provision of electricity or water. Electricity and water are both limited resources. The transfer of electricity and water are limited both by supply of the good and size of the pipe. Data over the Internet however, is unlimited. While the size of the pipe may be constrained, the amount of data is not. A more appropriate analogy is a highway, in this case an information highway."

This point could not be more wrong. You cannot transfer an infinite amount of information through a system with finite transfer limits. Not in a set period of time. This is why all data throughput numbers are calculated against time, in bits per second. A 14 kbps modem is capable of 14 thousand bits in one second. A 10 Gbps network link is capable of 10 billion bits in one second. So a system that can only transfer x bits of data per second can only transfer 60x bits of data in one minute, 300x bits of data in 5 minutes and so on. These are all finite numbers within a given time interval. The only way to achieve infinite data throughput is to allow for an infinite of time. Well, if you can wait forever, then a 14 kbps modem will meet your infinite data needs as well as a 10 Gbps network connection will.

Given that their initial premise is flawed, I shouldn't have to continue but I will.

Point 12 and 13: Further, we would draw attention to the Commission’s key principle of transparency. Is there Internet congestion? Where, when? Congestion should not be taken on faith from incumbents.

In the digital age, transparent information on Internet traffic should not be difficult to reveal. Only with open data and transparency can markets determine where best to invest in new capacity to reduce congestion, if it exists, and to maximize return on investment.

This is a red herring argument. Who says it is easy? Just because it is the "digital age", it doesn't mean things are easy. In this advanced age, where's my flying car, my brain to machine interface?

Secondly, what is meant by "where". Telling us a router in Toronto was congested doesn't mean anything, telling us it was router A123FGX123 on circuit YB2134 doesn't tell us squat either. The information of "where" really only matters to the network engineers that understand the blueprint and technologies of the network. And lastly, these reports of congestion are generated by equipment. The reports can be faked. The only way to not "take on faith" the information provided by incumbents is to allow third parties to monitor the internal network resources of the incumbents. That is a big kettle of fish to get involved in, and is not "easy" or "simple"

Point 15: Open competition means innovation and choice. Canada’s wholesale providers are indeed paying for their use and they should be allowed the freedom to distribute their traffic appropriately. If they want to offer high download cap packages, this is the nature of markets, competition and innovation.

Incumbents charging smaller ISPs Usage Based Billing rates does not prevent those ISPs from offering "unlimited" or high download cap packages. The costs to the ISPs may now be higher, but that is purely a cost of doing business. The government should not be in the business of negotiating supply costs on behalf of private companies.

Point 16.1: We agree with the CRTC that when congestion occurs an ISP’s first response should always be to invest in more network capacity;

Another pie-in-the-sky populist sentiment ideal. First, it is physically impossible to build a network that will not experience congestion. Again: all networks have finite limits (thousands of them, as a matter of fact). When those limits are measured against the totality of the data available to be transferred, they are tiny in comparison. So "upgrade first" is a never ending constant cycle of development and upgrades. Using the metaphor of a highway, think of it as being constantly widened every god-damned month. It would be a nightmare. In any event, these upgrade costs will be transferred back to the consumer at some point (they do have to be paid for!) What this argument also hides is that the onus is on the incumbents to upgrade, none of this is on the individual ISPs. They are users only. So in effect, this submission is telling the incumbents they cannot increase the rates charged to their customers, but they must invest in upgrades. Companies have to be able to recoup costs, otherwise they go bankrupt. This submission is forbidding the recouping of costs. (And no, there is no issue of fairness here. Yes, the incumbents may be, and probably are, using their monopoly position to dictate exorbitant rates. But forbidding charging their customers while mandating investment is not a solution. Well it is a solution, the same way a bullet to the head cures brain cancer.)

Points 16.3 and 16.4 get into fairness of charging customers. This argument states that averaging costs across all customers is unfair. Only the customers that are using data or causing problems should be charged. So the Liberal Party is advocating Usage Based Billing based on who uses what and where it is used......

I am going to end this now. This submission is complete shit. If the CRTC had any sense, they would feed it through a shredder publicly, in front of the Liberal Party when submitted.

Look, there are only two real solutions to the incumbent telecommunication fiasco we have here in Canada. Either deregulate ownership completely and allow foreign companies to compete openly, or turn the incumbents into a not-for-profit Crown monopoly.

To use another metaphor, our telecom network is like a big creaky old wagon that we are forced to ride on together. We're bitching about the costs to maintain it, while eager competitors are standing in the wings, prevented by law from building newer better wagons and competing for business. Like all of you upset about this UBB fiasco, I want a solution, but I want to get off the fucking wagon and choose my ride!!!!


Ken Breadner said...

Geez, for a while there I thought you were *for* UBB.
I hate to even comment in your field of expertise, but what really gets my goat are the multi-billion dollar profits that RoBellUs are making. To me, it seems like they could take a fraction of those and invest in upgrading the system...The other issue with UBB is the ridiculous markup. How much does it cost to transfer a gig of data? Does it cost more to transfer the 26th GB after you've already transferred 25? Am I being naive?

Catelli said...

I'm not anti or for UBB, anymore than I am anti or for flat rate billing. Business accounting practices that lead to pricing decisions are little bit beyond my ken. Though I can see merits to both...

As much as corporate profits gall us, there's nothing illegal about them. Therefore there's nothing regulatory that can be done about them. We can't choose how profits are reinvested (And they are at somepoint. That's capital money, so it is reinvested in the company somewhere later. It serves no ones interest, including shareholders for those profits to just sit in a bank account.) I'm not a big fan of capitalism, but I'm also learning when to stay out of its way, and when it becomes counter-productive to interfere. Which is why we're at this point, the government's already interfered too much already (foreign ownership rules for starters). So who says the media companies are not reinvesting and upgrading? It is equally likely that demand is outstripping supply. This is something I'm running into at work, our network needs are starting to outstrip the equipment available to purchase! We've deployed some bleeding edge technologies over the last few years, and I can see that translating to the communications sector, you have to run like hell just to stay in place.

And no, it doesn't cost any more to transfer 1 Gb or 100 Gb. Just like it doesn't cost anymore when there are 400 cars on the highway or 40,000. The highway when empty is the same highway when full. However maintenance, upkeep and other costs required to run the system can grow, especially if constant work is needed to repair and/or grow the system.

As I hope you understand by now, every network is over-subscribed. If everyone tried to use their maximum available bandwidth all at the same time, nothing would get through. So capacity planning for the future reflects the usage of the past. If people only use 10% of their available capacity per day, then the network is sized approproately with say a 1% annual growth rate. So in 10 years, people are using 20%. What kills things is when growth is on a curve, people are upgrading circuits and using them more heavily. It becomes a compound growth problem that is very hard to stay ahead of. And, how do you charge back the upgrade costs? Do you up everyone's base rates 25%, or do you start tacking on UBB costs to the top 15%? Is it fair that Grandma's circuit used for only e-mail and the occasional webchat with the Grand-kids goes from $30 a month to $50, or should the guy downloading movies left right and center assume a larger portion of the costs?

Now as to the true cost of running a telco: we will never know how much waste there is until true competition on the whole network happens. When someone builds competing infrastructure, and says to the smaller ISPs: "PSSSST! We can sell you net access at 40% of the cost Bell is charging you, want to talk to us?"

Imagine if your store was the only grocery store in town, and it was enormously profitable as a result. But instead of competitors building their own stores, your store was forced to allocate 25% of its retail space to competitors. And then their customers started bitching about how much your store was charging in rental fees to its own competiotors setting up shop in its own stores! That's a ludicrous scenario isn't it? But that's our telecom sector!

Ken Breadner said...

Thank you for this. It's really hard to cut through (and not fall victim to) the anti-UBB hysteria. You have explained this in a way that makes sense...except that so much of the current system we have doesn't make any sense.

Catelli said...

NP! I get the anger, I really do. I'm trying to make everyone see the whole forest.

I just wish someone in our government would do the same instead of pandering to legitimate, but misdirected, anger.