The UK regulator, Ofcom, released its annual communications markets report yesterday and it is an interesting read [pdf, 4MB].

There is a wealth of information to be found there – not unlike the CRTC’s report. One table in particular caught my eye, especially in view of the attention being paid to Blackberry.

Apple’s iPhone is the most popular brand of smartphone, with a 32% share among adults. This is the brand of choice among ABC1s (37%) and is even higher among ABs alone (44%). But BlackBerry handsets have also taken a significant share of the market (24%) and are particularly popular among younger adults and teens (37% each). Female teens, in particular, appear to have a preference for BlackBerry handsets (44%). Anecdotal evidence suggests that this preference is driven by the BlackBerry messenger service (BBM) which offers a free alternative to texting (SMS). [click here for description of ABC1, AB, etc.]

What is driving UK teens to Blackberry? Would we not have expected this to be an iPhone or Android group? Is the UK market an outlier or is this phenomenon taking place in other geographies? Is it the Blackberry physical keyboard or BBM that is driving its success?

If usage isn’t curtailed, is an economic traffic management practice ineffective? If there is no congestion being experienced, should traffic management practices (economic or technical) be prohibited?

These are some of the questions raised in arguments submitted last Friday to the CRTC in its wholesale usage based billing review.

MTS Allstream cited evidence from Cogeco (that stated its traffic growth rate had not significantly been impacted by retail usage based billing) to state that

volume-based billing models does nothing to reduce the total amount of traffic on the network… In other words… volume-based billing is ineffective in achieving its stated objective of suppressing volume of data transferred in a month.

Is reduction in traffic really the objective, or are carriers that implement economic traffic management practices looking to align retail prices with the value of service received? If volumes of traffic aren’t being changed, is there a fundamental problem with the practice? What policy objective is being violated?

In a strange twist of logic, some argue that users are generally not seeing evidence of congestion, so traffic management practices are not necessary. Would these people be happier to have degraded service levels first – the way cities manage road expansion. As if we would be better off to have carriers use “just too late” provisioning practices, instead of “just in time.”

Perhaps the most difficult piece to follow was submitted by CAIP. As an aside, CAIP’s piece was submitted late “due to technical problems”. It is somewhat ironic that a division of the Canadian Advanced Technology Association would be subjected to technical problems causing a two day delay in filing. CAIP raised the issue of Canada’s wireless industry to discuss network investment.

If we use wireless telephony services as an example, it becomes apparent that the frequency of investment and the value of these investments are complimentary [sic].

I’m not sure where CAIP was going with this. Like internet services, the wireless retail business is forborne from price regulation, Canada’s wireless industry does not have mandated resale requirements. CAIP acknowledges limited regulatory obligations on wireless:

Similar to the evolution of wireline networks, this is occurring because the three major wireless incumbents in Canada are desperate to gain market share from each other. In the wireless realm the key mandatory regulatory obligation is that the incumbent carriers provide access to their essential tower and site facilities and out‐of‐territory roaming.

Was CAIP suggesting that there may not be a need to have any requirement for regulated wholesale access, except for services provided to other facilities based carriers?

Ryerson Professor Catherine Middleton submitted comments to the CRTC’s usage based billing review that shared some international experience.

Perhaps most helpful was her reference to Australia:

The Australian Competition & Consumer Commission suggests that effective competition is characterized by rivalry in pricing and services, requires low barriers to entry for new providers, and enables competitors to hold “a reasonably sustainable market position.”

However, citing the CRTC’s recently released Communications Monitoring Report, Paragraph 12 of her comments assert that that “non-incumbent /competitor ISPs” hold just 6% of the market. In fact, the CRTC’s report shows that the major telephone companies had just 37.6% share. The 6% figure ignores the 56.4% by cable companies which cannot be combined with telephone company share. Surely the 62% share held by non-telephone company industry participants would be considered to be “a reasonably sustainable market position.”

I noted some other statements that might have drawn some questions from the CRTC, had she participated in the oral proceeding. For example, the comments assert

Taking the availability of fibre-to-the-home services as an indicator of innovation, with less than 1% of broadband Internet subscribers using a fibre connection, Canada is not a leader in providing innovative services.

Personally, I would not have considered the use of FTTH technology as an indicator of innovation, but rather an indicator of investment. As the CRTC’s monitoring report indicates, about 75% of Canadians have access to internet services at greater than 50Mbps, and nearly 20% of Canadian households have access to more than 100Mbps services. Both of these are substantial increases over the previous year.

Further, a reliance on OECD data would obviously not get my support, but there is a perspective that simply does not agree with the CRTC’s own data:

with just 6% of the market, non-incumbent/competitor ISPs in Canada have had minimal impact on consumer choice, pricing and the development of innovative services.

The CRTC’s figures show substantial changes in pricing [see Table 5.3.3], network development [Table 5.3.5] and availability of alternative networks [Table 5.3.4]. One cannot simply ignore the competition between telephone companies and cable companies. The submission refers to DSL and DOCSIS 2.0 as current generation while considering the broadband capabilities beyond these technologies as next generation infrastructure.

As Canadian Internet service providers build out next generation networks, it is essential that the regulatory environment continue to promote competition and ensure that Canadians have a choice of differentiated, innovative service offerings. In order for this to happen, competitive ISPs must remain in the market. The conditions in place in the wholesale market today will influence the nature of competition in provision of next generation Internet services tomorrow. It is essential that the CRTC preserve the viability of competitive ISPs so that they may continue to challenge incumbents and encourage the development of better Internet services for all Canadians.

It seems to me that the evidence shows that facilities based ISPs have been investing in these “next generation” technologies on the basis of inter-platform competition, far more than the competitive influence by independent ISPs. If the premise of the argument – that competitive ISPs encourage development of better Internet – doesn’t hold, is it no longer essential for the CRTC to preserve their viability? Is it the role of a regulator to preserve competitors or competition?

Friday’s mail had final submissions for the CRTC’s review of usage based billing.

I had a chance to glance through some of them over the course of the past few days.

At the essence, the challenge for the CRTC is to strike the correct rates, as Primus states right up front. Regardless of the model chosen, if the rates are set appropriately, independent ISPs will be able to develop differentiated services to help provide consumers with competitive options. It appears that there is no longer an issue about whether there will be a form of usage sensitive wholesale pricing, but rather what form of usage based billing will be adopted. As Bell stated:

Investments continue to be the first and preferred approach to managing network capacity but it cannot realistically be the sole approach. Without some form of pricing model that ensures that those who “use the most, pay the most”, revenues necessary to cover the costs and make investments in the network have to be absorbed by all users. Under the current wholesale model, those who use the least must subsidize those who use the most, and that is not the fairest approach to billing for Internet use.

We have two models under consideration, each with slight variants being advocated by various proponents: Capacity based; and, Volume based.

While Primus suggests that the capacity based model is attractive because

• Capacity billing models rely on industry standard practices.
• Capacity billing models correctly link billing amounts to the impact that competitors have on the network. This correctly ensures that the competitor that has the most impact on the network pays the most. This also provides a direct incentive for competitors to manage the impact that they have on the network.
• Capacity billing models are transparent as both the competitor and the wholesale access service provider are able to measure capacity usage at the same time and place. This significantly lowers the likelihood of disputes and streamlines the process to resolve any disputes that do occur.
• Capacity billing models completely sever any relationship between wholesale and retail billing models.

However, as Bell has indicated, there is not necessarily a correlation between peak usage at the point of network interconnection (between the retail and wholesale ISPs) and the various portions of the shared network that may experience congestion.

95th percentile is currently used as a billing method with regard to point-to-point services such as network transit and enterprise connections. In such circumstances 95th percentile adequately represents usage of the point-of-interconnection.

Wholesale access is more of a point-to-multipoint type service, for which Bell argues that a volume based billing model is a better proxy. On the other hand, TELUS does not charge usage sensitive wholesale or retail rates nor has it proposed to do so. As such, TELUS has requested that the CRTC continue to allow wholesale services to be offered on a flat rate basis, should it decide to.

TELUS supports the general position that an ILEC or cable carrier should be able to charge volume-sensitive wholesale rates should it wish to do so. All networks are not alike. Different architectures, different levels of investment, different usage patterns and different life cycles necessarily mean that networks will experience congestion and usage driven costs differently. TELUS has observed that data usage has been growing at an accelerating pace amongst its Internet customer base, including those end-customers that are served by wholesale ISPs. As a result, TELUS (or any other network provider) might ultimately, in the near to medium term, be required to examine volume-sensitive pricing models to deal with the costs associated with managing this data growth.

In choosing between usage sensitive billing models, TELUS supports an aggregated volume based model, because it is relatively simple to implement, easy to understand and less costly. A submission on behalf of Rogers, Cogeco and Videotron agreed and discussed the risk of arbitrage that would be enabled by the 95th percentile model.

I’ll have more later in the week. I welcome your comments.

WiFi safety was back in the news thanks to a tweet from Green Party leader Elizabeth May:

It is very disturbing how quickly Wifi has moved into schools as it is children who are the most vulnerable.

Which set off a maelstrom of tweets from the extremes of the informed.

Some pointed to the World Health Organization advisory on mobile phones and had trouble discerning the differences. A key statement in the WHO advisory appeared to have escaped from people’s view:

The power (and hence the radiofrequency exposure to a user) falls off rapidly with increasing distance from the handset. A person using a mobile phone 30–40 cm away from their body – for example when text messaging, accessing the Internet, or using a “hands free” device – will therefore have a much lower exposure to radiofrequency fields than someone holding the handset against their head.

Most of us do not walk around with a WiFi device strapped to our heads. Our laptop computers are usually more than the 30-40 cm distance away, at least when I am typing on mine. The Ontario Medical Officer of Health has said WiFi is OK, citing Health Canada standards. Princeton University – a school with a reasonable reputation – did a review of the literature to develop a position statement for its staff and students to provide assurance that they were working in a safe environment.

What was the basis for whipping up anxiety over technology that is already built to meet national and international safety standards.

It is still July. I wasn’t ready for back-to-school special interest issues like this, despite the pre-release announcements from some of the phone companies. Next thing you know, the office supply stores will start those “most wonderful time of the year” jingles.

UPDATE: David Hillier provides the best link to sum up the WiFi issue.