It has been a few years since I last looked at the economics of broadband. What are the business case drivers and roadblocks that arise when pushing network upgrades into suburban and rural networks?

As the CRTC begins its review of wholesale internet access, the business case for network investment, and the economics of broadband expansion, should be significant considerations.

A few years ago, I wrote “The economics of broadband expansion”. In that piece, I looked at the economics of broadband expansion from an engineering economics perspective.

In considering the economics of broadband, look at a piece I wrote in 2016 (“FTTH business isn’t binary”) when Minister Navdeep Bains rejected an appeal of the CRTC’s 2015 wireline wholesale decision providing fibre access, but only on a disaggregated basis. That decision says, “the provision of aggregated services will no longer be mandated and will be phased out in conjunction with the implementation of a disaggregated service”.

In each of these, I explain that there is a boundary defined as where the business case for network upgrades goes negative. We know, based on the design of all of the government broadband programs, that there are communities or regions that simply cannot generate sufficient revenues to justify the cost of building broadband facilities. A government subsidy intends to provide just enough money to make the business case go positive. That is, the subsidy provides a reasonable opportunity for a return on the private capital invested.

No subsidy is generally required in areas with an urban population density, other than remote areas that may need backbone transport. So, we have urban areas that have no need for a subsidy because the business case is positive. There are many rural areas that need a subsidy because the business case is negative. A kind of contour map can be drawn defined by where the business case for fibre construction is barely at break even.

That was the concept more fully discussed in “The economics of broadband expansion”.

But let’s go back and look at the ‘urban’ side of the boundary line. That boundary is defined as precisely where the business case goes from positive to negative. Homes on the boundary effectively have a net present value (NPV) of revenues less costs of zero. The boundary defines a form of digital divide.

What happens if wholesale and retail revenue assumptions change? If revenues increase, the boundary pushes outward. More homes (on the ‘rural’ side of the boundary) would potentially now have a positive business case. On the other hand, if revenues decrease, the boundary contracts and the business case is no longer positive for as many homes.
I wrote about the CRTC’s latest review of wholesale internet access a couple of weeks ago. In its new review, I mentioned that Bank of America has said “the key will be at what rates.”

Why?

A business case looks at whether the inward cash flows from incremental revenues are sufficient to cover the incremental costs of building broadband facilities. The revenue forecasts clearly change under the CRTC’s preliminary view of the wholesale access model.

The Commission invites comments on several issues, including its preliminary views that (i) the provision of aggregated wholesale HSA services should be mandated; (ii) access to fibre-to-the-premises (FTTP) facilities should be provided over these services; and (iii) the provision of FTTP facilities over aggregated wholesale HSA services should be mandated on a temporary and expedited basis, until the Commission reaches a decision as to whether such access is to be provided indefinitely.

Despite massive levels of capital investment by Canada’s facilities-based carriers, a lot of areas still do not have access to gigabit speeds. As seen in the CRTC’s data collection, 1 in 8 urban households in Canada lack access to ultra high speed broadband.

If the CRTC doesn’t get the wholesale rates right, the boundary for positive business cases will shrink and more households will find themselves on the wrong side of digital divide for next generation networks.

That would be a failure in broadband regulation.

A very brief post today in order to highlight a new report from GSMA, “Socio-Economic Benefits of 5G: The importance of low-band spectrum” [pdf, 8.9MB].

The report was released last week as GSMA set out its vision for the ITU’s upcoming World Radiocommunications Conference (WRC2023), taking place at the end of the year in Dubai.

GSMA says the importance of low-band spectrum is because of the propagation characteristics of such frequencies. Low-band spectrum is particularly well suited for providing coverage in rural and remote areas. In addition such frequencies have better in-building penetration, providing ‘deep’ indoor coverage as well as capacity in urban areas.

Low-band spectrum is a driver of digital equality, reducing the gap between urban and rural areas and delivering affordable connectivity. Without sufficient low-band spectrum, the digital divide is likely to widen, and those living in rural areas will be excluded from the latest digital technologies.

The report has an interesting comparison of indoor 5G signal strength in the largest cities in Australia, Canada and Japan. Look for Figure 12 in the report.

As WRC-23 approaches, we will likely be hearing more about spectrum policy.

A communications failure earlier today at Fredericton International Airport was caused by thieves stealing copper cable.

Bell has experienced 60 such incidents in the past 12 months just in New Brunswick. Although a news report says 4 people were arrested last November for copper theft, the damage to critical infrastructure is continuing.

New Brunswick Power and individual homes have also been victimized by theft of power cables and copper pipes. A man died in 2019 after breaking into a power substation in Bathurst, NB.

Last October, the New Brunswick legislature amended its Salvage Dealers Licensing Act, so that scrap yards are no longer able to pay cash and are required to check government issued identification.

Bell is asking provincial and federal governments to increase fines and make changes to the Criminal Code of Canada saying such measures are necessary to improve the resiliency of Canada’s telecommunications networks. Addressing the scourge of thieves stealing copper cable was identified as a top priority in the recent report [pdf, 475KB] from the Canadian Telecommunications Network Resiliency Working Group

This isn’t just a problem for telecom service providers, as we saw with the impact on flights in and out of Fredericton today. Stealing copper cable from power and communications networks can be a matter of public safety, impacting access to emergency service bureaus, hospitals and first responders. And, it cost at least one would-be thief his life.

As Bell CEO Mirko Bibic wrote, urgent action on the part of government is needed as part of protecting Canada’s critical infrastructure.

As promised, here is further look at the 2022 edition of ISED’s telecom price study report [pdf version, 1.8MB].

In its press release, ISED claims the government “continues to deliver on more affordable telecom services”.

The report is clear, Canada’s wireless prices declined by an average of 2.6% across all levels, with declines up to 16% for the largest data plans in 2022. For home Internet, prices declined or were stable, an 11% decline was recorded for mid-range plans. The report also shows that regional competitors are offering prices up to 39% lower than the major national service providers.

I’m not convinced the telecom price study report, or the press release for that matter, is so clear. That paragraph starts with a sentence about wireless plans, then a line about home internet.

To which service does that third sentence refer? “Regional competitors are offering prices up to 39% lower than the major national service providers.” All of the facilities-based internet service providers are regional; the national providers are the wholesale based ones. You have to search the report to find Table 3, to find that the 39% refers to Freedom Mobile’s Level 7 price plan in Ontario. Level 7 is defined as “unlimited nationwide talk and text along with 20-49 GB of data.”

The report also indicates that “Average prices in Quebec tend to be among the lowest in the country.” This has been a highlight driving a number of policies out of Ottawa. Yes, prices have been lower in Quebec but do we actually understand the reasons? Is there enough focus on what is driving lower prices in some parts of the country?

Are lower prices in Quebec actually due to the presence of Videotron as a “disruptor”? Or, are lower prices driven by lower rates of adoption and attempts by service providers to stimulate demand in a province with lower average income levels?

Let’s look at what the press release says is “clear”. “Canada’s wireless prices declined by an average of 2.6% across all levels, with declines up to 16% for the largest data plans in 2022.” I’m not sure the data in the report is really that clear.

The source for this line appears to be this chart (found on page 10 of the pdf version):

As can be seen readily, the average price change (-2.6%) is heavily skewed by the nearly 13% price increase in Level 2 pricing. Of course, very few people choose to subscribe to such a plan any more. Indeed, the report itself acknowledges “very few providers currently have a stand-alone 1 GB plan. Only Virgin offers this service across the country and only one regional provider (SaskTel) offers this type of plan.”

Also, recall that the mid-range plans had already dropped by more than 25% prior to the start of last year. That helps explains why one might expect modest price declines for those price baskets.

Although the press release says “The report is clear, Canada’s wireless prices declined by an average of 2.6% across all levels”, what is actually clear is that prices declined significantly more if you exclude the obvious outlier.

As with other international price comparisons, look at the data with a critical eye. As a test, see if you can find the typo at the bottom of page 69 of the pdf version (Table A3.1).

The missing caveats from the older editions of ISED’s telecom price study should be returned. As I wrote earlier this week, “Prices in Canada and international jurisdictions are driven by a complex mix of a number of factors: cost of service, competitive positioning, technological advances, consumer behaviour and regulatory frameworks.”

In the absence of such notes to the reader, is it fair to describe any price study as “clear”?

A week and a half ago, ISED released the latest edition of its series of telecom price studies. I’m going to look at that report over the course of a few posts.

It’s a real challenge to create meaningful international telecom price studies.

Remember when two dozen leading economists and academics said, “The Rewheel story is easy to understand. It is also completely wrong.” and “Rewheel’s rankings are of no value in comparing prices and assessing the level of competition in wireless markets.” Rewheel’s reports were characterized by ICLE as “a careless mish-mash of data points from which no reliable conclusions can be drawn.” Last week’s report [pdf, 1.2 MB] lives up to that billing.

I looked at a couple well publicized international telecom price studies about a year and a half ago. In that post, I write of my frustration with “the misinformation from pseudo-statistical studies being circulated with viral velocity”. I pointed out what should be easy to detect flaws with the methodology being used by Cable.co.uk.

Typical problems with telecom price studies arise from overly simplistic examination of the different countries. While most studies adjust for currency variations, very few make adjustments for PPP (purchasing power parity). If consumers are earning 80% less in one country, it doesn’t help for them to pay 25% less for their digital connections.

Fewer still account for variances in quality of the products and services, such as speeds, coverage, costs of building networks. That can be like comparing prices for bicycles and motorcycles. A recent PwC study [pdf, 660 KB] compared Canada to the rest of the G7 plus Australia. Canadian carriers invest almost double the amount capital measured on a per subscriber basis ($168 vs $87), with capital intensity 35% more (19% versus 14%).

There are hundreds (or thousands) of price plans available in each country. It is virtually impossible for telecom price studies to look at which plans are the most popular in each market. And then, how would a study start to compare those to the plans in other countries? Arithmetic averages (means or medians) are somewhat meaningless. Are the plans that most consumers are buying are weighted more heavily than those on extreme ends of the menu? In countries with 150 to 200% mobile penetration rates, does the study account for people paying multiple bills?

Let’s consider the telecom price study released earlier this month by Innovation, Science and Economic Development: “Price Comparisons of Wireline, Wireless and Internet Services in Canada and with Foreign Jurisdictions: 2022 Edition” [pdf version, 1.8MB]. Wall Communications prepared the report for ISED.

As in some other recent years, the 2022 edition is missing a section on caveats to the interpretation of the findings. Those notes used to be an important part of the study. For example, in 2016, the ISED study included a page of notes, including these two paragraphs:

Prices in Canada and international jurisdictions are driven by a complex mix of a number of factors: cost of service, competitive positioning, technological advances, consumer behaviour and regulatory frameworks. As wireless technology is constantly improving and consumers demand ever more bandwidth and data caps, service providers are constantly increasing features. In the Study, these changes are reflected by the need to regularly update the definition of service baskets. Hence, price increases in those baskets may in part, simply reflect better service levels offered to consumers.

This Study did not take into account the network technologies deployed in the networks nor the speed or quality of service of those networks. Finally, this Study did not account for any cost of service or socio-economic factors that may be relevant for price differences across different domestic and international jurisdictions. Thus, factors such as population density, terrain and climate have significant impacts on the cost of service. Similarly, socio-economic factors such as affordability indicators (i.e. mobile prices in relation to disposable income), number of handsets per subscriber, number of minutes of usage per subscriber and other factors were not within the scope of this Study.

The 2022 edition includes a few caveats in its Introduction, but it would benefit from a separate “reader’s notes” section.

I’ll look at the results of ISED’s 2022 price report in another post later this week.