It’s time to look at splitting up Canadian telecoms

04 Jun

Chopping_BoardSo Canadians finally have a set of rules that will shield them from the worst practices of cellphone carriers - three-year contracts and exorbitant roaming rates among them - thanks to the Canadian Radio-television and Telecommunications Commission and its new Wireless Code of Conduct. But will the regulator’s manifesto result in lower monthly bills? In some cases, perhaps, but in others, no.

The code effectively makes three-year contracts moot by requiring carriers to divide up the cost of customers’ phone subsidies over a maximum of 24 months. If a customer wants to cancel early within that time frame, carriers are only allowed to recoup whatever subsidy is left on the device and nothing more. They’re still free to offer three-year contracts, but with customers able to walk away after 24 months with zero charges, there’s no point.

The rules also cap roaming fees at $100 and data overage charges at $50, unless the customer expressly agrees to more in either case. Both measures should prevent those horror stories where subscribers have been saddled with bills in the thousands of dollars. Carriers will also be required to unlock subsidized phones within 90 days and those fully paid for up front immediately. It’s all good stuff.

Carrier reaction will likely take the form of higher up-front costs on phones. Today’s $179 iPhone 5 on a three-year contract might quickly turn into a $400 device on a two-year deal. Instinctively that’s not good, but in the long run it’s okay because it will finally make prices - both on devices and services - comparable to other countries, meaning it’ll be easier to see if Canadians are indeed overpaying. I’ll have more to say on this tomorrow.

Over all, the code won’t make much of a difference, which is why analysts are expecting wireless bills - and profits - to climb higher. Canada is actually one of the few countries where that’s the expectation; in most other jurisdictions, the momentum on both measures is downward.

The onus is now on the federal government to fix the bigger issue, which is a lack of competition between big players. There’s little doubt that Ottawa’s effort to inject more competition into the market five years ago is teetering on failure. New entrant Mobilicity is being acquired by Telus, a big chunk of wireless spectrum reserved for new players is about to be transferred from Shaw and Quebecor to Rogers, and the other small independent players - Wind and Public Mobile - are reportedly running out of cash. Analysts are probably correct in saying that the past few years of somewhat lower bills were just a hiccup - Canadian wireless companies are on the verge of resuming business as usual. In that case, let’s get the pool started: how long till the system access fee makes a return?

The debate now turns to the larger issue. Can the government salvage its effort to drive more competition into the market and therefore lower bills, or does it cut its losses and effectively tell consumers, “Sorry folks, you’re out of luck?”

Plenty of suggestions are being thrown around. University of Ottawa professor Michael Geist, writing in the Toronto Star, correctly asserts that when meaningful competition between market players is absent, regulation is necessary. He suggests stronger rules governing things like the sharing of cellphone towers, a full set-aside of spectrum for new companies in the upcoming spectrum auction, and the complete removal of foreign investment restrictions.

Those are all good ideas, but Geist also suggests one terrible one: regulations forcing third-party access to the wireless networks of Bell, Rogers and Telus. Such a scenario would lead to the creation of a swath of companies known as Mobile Virtual Network Operators (MVNOs), who effectively buy discounted minutes and data from network owners, then resell them under their own banner.

Canada already has a number of these, with the likes of President’s Choice and 7-11 currently in voluntarily negotiated agreements with the big carriers. MVNOs are also enjoying something of a renaissance in the United States.

The thing is, MVNOs fail far more often than they succeed - Amp’d, which piggy-backed on Telus’s network, folded after only four months while Virgin was eventually swallowed by Bell - simply because their existence is predicated on the whims of their partner. Absent regulations, network owners get to call all the shots, from the rates MVNOs can effectively charge to the types of phones they’re allowed to sell. That makes it hard for the virtual operators to come up with any sort of viable, long-term business plan.

Instituting regulations to govern all that might be just as bad, if the wired internet side of things is any indication. As is painfully apparent to anyone who pays attention to this wholesale realm, it’s just one long tennis match in front of the CRTC, with independent operators complaining about mistreatment from big network owners in one form or another. Throttling and usage-based billing are two of the more recent examples.

It might all be worth it if those wholesale players were making some kind of a difference, but they’re not. Despite all the years of back and forth between the likes of Bell and smaller companies such as Teksavvy, smaller internet providers account for only about 6 per cent of the market. They generally provide better-priced plans, but ultimately such ISPs aren’t doing much to discipline bigger players or change things for the better. Given that, a similar scheme in wireless hardly seems worth the trouble.

It’s therefore time for government and regulators to revisit a more serious measure: the forced splitting of companies and the networks they own. Known as operational or structural separation, such a set-up usually involves big telecom companies carving off into a separate organization the parts of their respective companies that are devoted to running networks. That unit then sells access to the network to all comers, including the parent company, on a completely equal basis.

The network business is thus the ultimate wholesaler, with strong, institutionalized business incentives to maintain its neutrality (i.e. executives can get pay bonuses based on the total number of customers they sign up).

The concept has been effective in numerous infrastructure-based sectors that are inclined to natural monopolies, including railroads, gas and electricity. The logic has been pretty simple: in the cases of these industries, it’s simply too expensive, impractical or undesirable for multiple parties to build infrastructure. It simply makes much more sense for them to all share the same base on equitable terms.

Whether or not such separation is a good idea for telecom networks is still an open question, but there is mounting evidence that it is. The pioneering example is BT in the United Kingdom. The company initially fought separation tooth and nail, but eventually decided to do so voluntarily before the regulator forced the matter. In 2006, the company set up Openreach - a network-governing unit that would sell broadband access to anyone who wanted it.

Openreach’s success can be measured in a number of ways. Despite its initial reluctance, the move has been a boon for BT, which has seen its stock double since the launch.

Critics also routinely say that mandated wholesale arrangements discourage network owners from investing in upgrades, which is indeed what we’ve heard in just about every dispute between big companies and small ISPs in Canada. However, the Organization for Economic Co-operation and Development studied structural separation (links to PDF) across various industries several years ago and didn’t find that to be the case:

Where an infrastructure owner is subject to mandatory access requirements, it may choose to refrain from developing additional capacity on its network, even in the face of considerable demand, in order to prevent its downstream competitors from gaining access to the infrastructure necessary to supply the downstream market. Alternatively, as has also been an issue in the context of the switch to smart electricity grids, where the transmission system owner remains part of a vertically integrated firm, it has an incentive to implement network upgrades in a manner that excludes third parties from the competitive segments. In such circumstances, structural separation is likely to improve the infrastructure owner/operator‘s incentives to investment in the facility, or to do so in a manner that facilitates competition in non-monopoly sectors.

That’s exactly what has happened with Openreach. The BT unit is currently spending $4 billion on fibre upgrades in response to strong demand from wholesale customers, and is planning on being done 18 months ahead of time.

Most importantly among all this, broadband choice in the U.K. has exploded. Consumers have a huge field of providers to choose from, some of whom are offering insane prices. Supermarket chain Tesco, for one, is selling broadband access with unlimited usage for $3 a month. That’s as alien a reality in Canada as a winter without bone-chilling temperatures.

There is at least one clear downside to the separated wholesale system in that it creates a gap between the network owner and the end user. That means if something goes wrong with the service, there can be delays in getting things fixed, which leads to customer frustration. It’s happening with Openreach and it’s also happening with wholesale ISPs in Canada. It’s definitely one aspect of the scheme that needs to be further developed.

That hiccup notwithstanding, Openreach seems to be accomplishing the objectives of all U.K. stakeholders, from government and regulators to corporations to consumers: better returns for BT, more investment for the country, cheaper prices and better service for consumers.

The CRTC studied the idea of structural separation way back in the early nineties before ultimately rejecting it in 1994, opting to go with a less-stringent wholesale regime instead. Given the string of failures to prevent continually rising bills in both wired and wireless services, it’s time to dust the idea off and give it another strong look.

About these ads

12 responses to “It’s time to look at splitting up Canadian telecoms

  1. Marc Venot

    June 4, 2013 at 12:27 am

    The CRTC has patched the worst abuses.
    Cellphone is a large cash cow for governments and since it’s somehow voluntary like gambling, it’s not as bad as taxes.

  2. globeman

    June 4, 2013 at 12:33 am

    split up the cable companies. they have the superior wired network (more bandwidth). The technological limitations of copper twisted pair makes it difficult for telephone companies to wield much pricing power.

  3. Ben Klass

    June 4, 2013 at 2:42 am

    Note should be taken of the NBN in Australia as well - the government outright bought a fibre backbone from one of their nation’s largest providers - at a reasonable cost of $35 billion AUS, and allows ISPs to provide service by buying access to the fibre net at a regulated rate.

    Canada came close to a similar scheme - nationalization, that is - in 1905, with the Bell. Public outrage reached a fever pitch after 20+ years of a policy of regulatory abstention, which was abused (price gouging, anti-competitive behavior, collusion with other industries, most notably the railroads) to no end by Bell. Sound familiar?

  4. Infostack

    June 4, 2013 at 6:44 am

    Wholesale worked and works well when there is transparency and the market or regulator have a view to true marginal cost at every layer and boundary point. Which, of course they do not because their models and theory are flawed, based on a 100 year old precept embodied in the Kingsbury Commitment. With the right model and data, wholesale could conceivably work just by enforcing open/equal access in layer and applying it across the board.

    Likewise structural separation should be applied universally, meaning to all incumbents and newcomers alike; wired and wireless, video and voice and data. Call it the Big Bang 2. It’s necessary anyway because the horizontal digital forces of OTT, WebRTC, SON/hetnets (self-organizing networks, offload, etc..), and SDN/openflow (software defined networks driven by data-center and transport layer scaling) are crashing against the vertically integrated service provider model across all applications and market segments. So best do it quickly than watch a slow, expensive and wasteful erosion that merely impedes economic growth and the digital restructuring and regeneration of all socio-political and economic institutions.

    The biggest problem is the unions, who will fight it tooth and nail, even though it will be a highly generative solution for them directly in terms of increased investment (opex and capex) in every layer and indirectly via economic growth. However it will break their vertical monopoly. So any structural separation case will need to address the overall labor impact to counter any resistance on that front.

    The economics and approaches behind the internet and 802.11 are the best models to follow, although the former still refuse to accept that we need balanced settlements in the middle layers (known as 2-sided revenue models in the monopoly world) that clear supply and demand both vertically and horizontal in the stack to create push/pull dynamics for rapid service creation/spread, infrastructure investment and centralized procurement, whereby core players/solutions pay for or subsidize edge access. In a competitive model, these centralized payments would drive price/bit even lower, so it is not a 2-sided approach taking, rather a generative process that helps supply continue follow moore’s and metcalfe’s laws efficiently and generate ROI. We saw this process occur in voice, data and wireless in the 1980s-90s in the US where pricing in each instance dropped 99% over 10 years. Bill and keep simply does not work and results in inefficient balkanization.

    Surprising why Google is not pursuing this strategy in the US, or at least hinting at going that way. If you read CNET’s recent review they clearly say that the video headend was the most challenging and expensive aspect of the undertaking, once they got the govt to ease barriers to rapid build out of the backbone and into the fiberhoods. Here are the flaws, which should be getting addressed, but as best I can tell, no one has done so, since very few have a “horizontal” approach/understanding:
    1) closed access in layer 1, no resale to 3rd parties, no ability to resell wifi at the edge to neighbors
    2) 1x purchase event; SoL if you don’t hook up when their truck comes to your neighborhood
    3) no indication of licensing the fiber buildout model regionally/nationally so that others (muni’s, incumbent LECs/cables, new cos, etc…) can invest in it and google can amortize it’s hardware investment and scale both it’s WAN transport and possibly future optical switching
    4) no scaling of the video investment by licensing to 3rd parties nationwide to begin to disrupt the bundled cable model
    5) no apparent amortization from wireless (be it backhaul for 4G or high-capacity 802.11 offload ubiquitously)
    6) no commercial access and services (why not go and disrupt the enterprise market?)
    7) no centralized procurement, such as corporate VPN for 50 mbs CIR for 100 or 1000 teleworkers to enable HD collaborative conferencing (can be applied to telemedicine, education,etc…)

    If we add all of these “horizontal” elements to the model, it scales well beyond the $70 for 1gbs, which is consistent with ongoing moore’s and metcalfe’s laws which drive the cost/bit down 40-50% annually when combined. Right now it looks like Google just wants to build an alternative, albeit low-priced, vertically integrated look-alike model to the incumbents.

  5. Anthony Reimer

    June 4, 2013 at 12:04 pm

    The recent Canadian example of structural separation has been with utilities (natural gas, electricity) in Alberta. I do not think it is overstating the matter to call that change an abject failure for consumers (I know I have never signed a contract with a utilities retailer). While the argument is interesting, I do not believe it is compelling for the Canadian market. I really don’t see a palatable solution other than stronger regulation and I don’t see that happening without a change in government.

    • Infostack

      June 4, 2013 at 12:33 pm

      @AnthonyReimer 2-way communications networks are fundamentally different than the “other” natural monopolies. The utility model will begin to look a little bit more 2-way when smart-grid enables, distributed co-generation, storage, energy management and electric vehicle charging models. However, we have quite a bit of proof with 3 waves of digitization fostered by varying forms of equal access in the US in the 1980-90s that structural separation would and will work. It’s the right approach, data framework and policy oversight that will be necessary in the lower, middle and upper layers that will make it work. Most government oversight and intervention has been an abject failure in communication networks dating back 100 years to the farcical and fraudulent Kingsbury Commitment. I could make a case how the latter set us back 50 years in the digital information revolution begun 170 years ago.

  6. Serge

    June 4, 2013 at 1:10 pm

    Probably worth mentioning that Tesco’s $3 broadband requires you also to buy the underlying line for $23.50 on a 12-month contract, so the real cheapest-in-the-UK price is $26.50. That’s still pretty good for unlimited subject to fair use policy, and they throw in free evenings and weekends VoIP calling. Teksavvy’s cheapest is about $35 (with unlimited VoIP, for comparable) on no contract.

    Why haven’t major Canadian retailers yet gotten into virtual ISPs in the way that Tesco has, and the way that Canadian retailers have in terms of long distance cards and, more recently, mobile? Do we think it’s coming? In fact, would white-labelling their broadband service for a Canadian Tire or a Loblaws not be a growth path for the Teksavvys of the world, who already have the expertise?

    • Peter Nowak

      June 4, 2013 at 3:50 pm

      The answer to that one is pretty simple: it would be too much of a pain in the butt for retailers to get into the ISP business. Who in their right mind wants to be part of the constant squabbling over rates and network access that is the wholesale business in Canada? Of course, that would change dramatically with a structurally separate system like the UK’s.

      • Marc

        June 4, 2013 at 5:47 pm

        A carrot and a stick… Make wholesale work.. Or else! Many recent changes have already been made also.. Transparency to costing may go a long way to help for example.. And above all, an engaged CRTC that attempts to chase down the truth…

  7. Mike

    June 5, 2013 at 7:13 am

    Why dick around with half-measures? It’s an essential infrastructure - nationalize it.

    • Michael Elling (@Infostack)

      June 5, 2013 at 7:27 am

      This is the worst idea.

      You want a body that can’t make any type of risky decision and provides half measures at beest to build your layer 1? You want them to maintain and upgrade it the way they work on our horrible roads?

      Do you know how rapidly technology is changing in layers 1 and 2 on both the wired and wireless side? The only “real” issue other than ensuring equal access and keeping an eye on any potential monopoly choke points across all layers and boundary points, is to determine/arbitrate physical rights of way issues. As for frequency, 802.11 has proved the value of sharing spectrum, and if specific channel licenses are granted they must be open.

      Let the markets decide the tradeoff between transport and switching and edge and core processing to handle exploding and diverging demand.

  8. bradfonseca

    June 5, 2013 at 8:33 am

    Reblogged this on The Emporium of Lost Thoughts and commented:
    Excellent discussion of the topic of Structural Separation and how it could help improve access to and quality of broadband and wireless services at a lower cost.


Get every new post delivered to your Inbox.

Join 267 other followers

%d bloggers like this: