Friday
Oct262012

Comparing Canada's Broadband Performance

A country’s performance in delivering broadband internet services to its citizens is widely recognized as way to gauge its economic and social well-being.  There are numerous indicators and sources of broadband performance.  Two that are widely cited are adoption, measured as the percentage of households or population with a connection, and the speed of the connection. 

Adoption rates generally reflect both the availability of a connection, and value of service as perceived by consumers.  Adoption of fixed broadband internet services can be measured as a percentage of households, since typically one household will share the connection.  Adoption of wireless broadband internet services, in contrast, is measured as a percentage of the population since most connections are dedicated to an individual smartphone.

The speed of the broadband connection is an indicator of the quality of the service.  Higher speed connections support more bandwidth-intensive applications, for example, real-time video streaming and video chat services.  The speed experienced by end-users, rather than the advertised speed of the service that they subscribe to, is a reliable basis for comparison.

The following charts compare Canada’s broadband performance with other G7 countries in terms of adoption and speed during the past five years.  These indicators are for fixed broadband connections. 

Adoption of broadband internet is taken from the OECD’s latest series up to the end of 2011.  The OECD measures subscriptions relative to population, rather than households.  As a result, it will somewhat understate household adoption in some countries, including Canada, where household size is higher than in other G7 countries, notably France, Germany and the United Kingdom, according to the OECD.  The OECD defines a broadband connection as one that is at least 256 kbps.

 

 

Canada’s adoption rate in terms of population ranks below France, Germany and the United Kingdom.  In the case of the latter two countries, Canada is sufficiently close that if adoption of fixed broadband service was measured in terms of households, it would rank higher, as indicated in Figure 6.1.5 of the CRTC’s latest Communications Monitoring Report.

While Canada is ahead of Japan in terms of fixed broadband connections, it should be noted that Japan has a very high level of mobile broadband connections.  It is possible that some amount of substitution of mobile for fixed connections has occurred, which would boost Japan’s overall broadband penetration.

The speed of internet connection is taken from data reported by Akamai as part of its “State of the Internet” analysis.  In addition to having recently released data for the second quarter of 2012, Akamai also provides a data visualization tool that allows users to review historical data on various indicators beginning with the third quarter of 2007.  Service speeds are based on achieved, rather than advertised, speeds.

In terms of the overall average connection speed, Canada and the United States continue to battle it out for second place among G7 countries at 6.4 to 6.5 Mbps, well behind Japan at 10.5 Mbps, as of the second quarter of 2012.  Germany and the United Kingdom have also achieved speeds similar to each other, at 5.5 to 5.6 Mbps, while France averaged 4.5 Mbps and Italy 3.9 Mbps.

Another indicator of broadband performance is the percentage of connections that exceed a certain minimum speed.  In 2012, Akamai redefined broadband connections as those providing at least 4 Mbps downstream.

 

Again, Japan leads the G7 with almost 74% of fixed broadband connections above 4 Mbps.  Canada has been catching up, closing in on 69% of connections as of the second quarter of 2012.  The remaining G7 countries are further behind, with the U.S. and the United Kingdom striving to get to 60%, France at less than 50% and Italy less than 30%.

Canada’s proportion of broadband connections above 4 Mbps was ranked sixth highest among all countries tracked by Akamai in the second quarter of 2012, up from an eighth place ranking in the previous quarter. 

Canada does not compare as well in terms of connections above 10 Mbps, with only 11% achieving that level.  In Japan, 37% of connections are faster than 10 Mbps, and 16% in the United States.  The other G7 countries have fewer than 10% of connections surpassing 10 Mbps.  While Canada ranked third among G7 countries, it would rank 15th among all countries tracked by Akamai.  By comparison, Japan ranked second behind South Korea while the United States was seventh.

Monday
Oct222012

Television Market Shares

The CRTC rejected an application by BCE to acquire Astral Media in Broadcasting Decision 2012-574.  One element of that decision that is getting considerable attention is how to calculate the market share in television programming that BCE would control if the acquisition had been approved. 

There are a number of options for calculating market shares, depending on what is selected as the unit of measurement, the definition of the overall market and the attribution of units to specific parties, in this case, BCE+Astral. 

Two leading contenders for calculating market share in the case of television programming are revenues and viewership among Canadians.  The CRTC’s decision considered both, but provided the most detail on the market share based on viewership.  The importance of the calculation also derives from the CRTC’s market share thresholds set out in its Diversity of Voices ruling (Broadcasting Public Notice 2008-4).  In that case, market shares are based on shares of viewership, measured on a national basis, separately for Anglophone and Francophone viewers.  

BCE claims that the CRTC got the calculation of viewership shares wrong because it excludes viewing to non-Canadian programming from the definition of the overall market.  In BCE’s view, non-Canadian television programming, such as CNN or A&E, competes with Canadian television programming for Canadian viewers, and viewing to such programming should be included in the definition of the overall market. 

The CRTC publishes its Communications Monitoring Report each year that includes statistics on the number and share of viewing hours by language and type of television programming service (e.g., Canadian conventional, specialty, etc.) and by corporate ownership of the television programming.  Some of these figures were also provided in the background information released by the CRTC with its decision to reject BCE’s purchase of Astral. 

These sources provide enough information to calculate viewership market shares using the same method as the CRTC and then adjusting it to use the definition of the overall market that includes viewing to non-Canadian television programming, as well as viewing to third language programming services.  The difference is mainly with the Anglophone viewing shares since Francophone viewers have very limited viewing to non-Canadian television programming.

The two charts provide a side-by-side comparison of the market shares for television viewing in the Anglophone market using the different definitions. 

 

Sources: CRTC Broadcasting Decision 2012-574 and Background Information to the accompanying press release, October 18, 2012; CRTC Communications Monitoring Report, 2012, Tables 4.3.3, 4.3.9.

Under either calculation, BCE+Astral would have a market share that exceeds 35%.    The 35% threshold was established in the CRTC’s Diversity of Voices as representing the level below which a company could expect its acquisition to be approved, “barring other policy concerns”, as noted at paragraph 87 of that ruling.  

BCE claims that, in rejecting its acquisition of Astral, the CRTC changed its approach to calculating market share from that relied on in its Diversity of Voices ruling.

The CRTC’s Diversity of Voices ruling did not describe the specific methodology to calculate the market shares for assessing proposed acquisitions against the thresholds.  Although it did provide some television market shares for larger broadcasters based on information available at that time.  CTVglobemedia (CTV), which BCE took control of and renamed Bell Media in 2011, was stated to have a 37.4% market share in 2007, as noted at paragraph 80 of the Diversity of Voices ruling. 

The CRTC revealed more information on its methodology for calculating the market share values in its Communications Monitoring Report, 2009.  The CRTC discussed its Diversity of Voices ruling at pages 117-118 of that Report and gave updated market shares for CTV and the other major broadcasters.  CTV’s share was reduced to 34% for the 2007 broadcast year, and measured as 33.4% in 2008.  These shares correspond to those presented in Table 4.3.10 of the CRTC’s 2009 Report. 

The information provided in the Report indicates that the CRTC’s thresholds for its Diversity of Voices defined the total market for viewership based on viewing to Canadian services only.  If viewing to non-Canadian services had been included in the definition of the overall market, CTV’s share would have been 24.4% in 2007 and 24% in 2008, based on the data provided in Tables 4.3.4 and 4.3.10 of the Report.  The gap between these market share values for CTV and the 37.4% provided in the CRTC’s Diversity of Voices ruling provides further confirmation of the methodology used. 

The CRTC has consistently calculated market shares for the purpose of its Diversity of Voices thresholds based on excluding viewing to non-Canadian services from the definition of the overall market.   This is the same method the CRTC used to calculate the market share for BCE+Astral in its most recent decision.

The above analysis and opinion reflect solely the views of Giganomics Consulting Inc.

Thursday
Aug252011

Wireless Substitution - Update

About a year ago, we looked at the state of wireless substitution, measured in terms of households that have no wireline telephone service and rely only on wireless service.  Earlier this year, Statistics Canada released some figures on wireline and wireless telephone use in 2010 based on its Residential Telephone Service Survey.  The size of the survey sample (more than 19,000 households) provides a greater degree of reliability, particularly when looking at smaller subsets of the population.  The following update is based on the Statistics Canada summary results released in April 2011, as well as more detailed survey data that is publicly available.*

The percentage of wireless only households in Canada had risen to 13.7% by the end of 2010, representing nearly 1.9 million households.  The incidence of wireless only is highest among households made up of younger adults (aged 18-34) and those who rent.  One-half of younger households were wireless only in 2010, up from 34% in 2008, while renters continued to account for one-third of all wireless only households.  It is also interesting to note that Vancouver and Victoria had the highest incidence of wireless only households in 2010, at 21.7% and 20.5%, respectively.  

An additional 1.8% of households reported having a wireline phone that was used only for business, computer or fax purposes.  This suggests that the number of households that rely on wireless only for residential purposes is more than 15%.

The growth in wireless only households in Canada was accompanied by a continued decline in the number of wireline connections.  As noted previously, local wireline connections peaked in Canada in 2006 at 21 million, based on CRTC statistics. As of 2010, 1 million wireline connections had disappeared, even as the number of households increased. The number of wireless subscriptions outnumbered wireline by 5.8 million.

Source: CRTC, Communications Monitoring Reports

The trend in wireless substitution is even stronger in the United States.  According to data published by the Centers for Disease Control and Prevention, 29.7% of all U.S. households were reported to be wireless only as of the close of 2010, up five percentage points from the previous year.  An additional 15.7% of U.S. households relied on their wireless phones for all or almost all of their calls.  These two groups of households account for more than 45% of all U.S. households. 

Canadian trends in wireless substitution appear to be following the same pattern as in the United States, with a delay of a few years.  

The accelerated shift to wireless that occurred south of the border in 2010 could have been due in part to economic pressures that may not occur to the same degree in Canada.  Even still, U.S. trends suggest Canadian wireless only households could reach 30% by 2014.  

The increasing functionality of wireless devices could drive wireless substitution at a faster pace.  Improvements in mobile networks and the growth in applications on smartphones are key contributors, as noted in our observation last year, and echoed in the CRTC’s second report on Navigating Convergence.

*This analysis is based on Statistics Canada Microdata file the Residential Telephone Service Survey which contains anonymized data collected in the 2010 Residential Telephone Service Survey. All computations on these microdata were prepared by Suzanne Blackwell and the responsibility for the use and interpretation of these data is entirely that of the author.

Thursday
Jun092011

Canada mobile data roaming fees - another sad story?

The OECD released data guaranteed to cause anxiety and anger among Canadian mobile users. This time it is targeted at mobile data roaming fees. The OECD press release included a chart that ranks Canada as the most expensive country to have your mobile data service from when travelling. The OECD called on regulators and policy makers take action to bring down data mobile roaming fees.  

The OECD report accompanying the press release has more information and it is worth taking a closer look.

First, the data given in the press release is a simple average of all the data rates that apply when roaming in each of the 34 OECD countries. If there are a few countries where data roaming rates are high, this will push up the average even if these are not frequent destinations for Canadians.  A proper comparison would have weighted the prices by relative volumes or value for each destination. This information was not available for the OECD report.

Most Canadian travellers visit only a handful of countries and the vast majority of our visits – more than 70% in 2009 – are to the United States. (Statistics Canada provides a list of the top 15 countries.)  A simple average of all the rates does not match this reality.

Second, the press release comparison is based on using a single MB. Data usage on most mobile devices is considerably higher.  For example, Cisco recently reported that data usage across all mobile devices in Canada averaged 4 Petabytes per month in 2010. This equates to 4 trillion MB per month. Compared to the 5.7 million mobile broadband subscribers in Canada, the average mobile broadband subscriber consumes more than 750 MB of data each month, or about 25 MB daily. Undoubtedly, only a fraction of that would be outside of Canada, but it gives an order of magnitude of typical daily use.

Third, the press release also focused on prices for a single user-initiated data session, within one day.  The travel information provided by Statistics Canada referenced above indicates that Canadian visits to other countries typically last at least one week.  A price comparison based on use within one session or day does not match reality. While the prices for Canada plans do not vary if the use occurs within one session or multiple sessions, there are significant differences in other countries (notably Japan).

In fact, the detailed OECD report included fees for data usage that allows us to overcome all the shortcomings of the press release comparison. There are prices for the least expensive destination - also the most popular destinations; for higher levels of usage (5 MB and 20 MB); and with usage spread over multiple sessions during a week or month. It is illuminating to consider the results of the price comparison based on these scenarios.

The following chart compares Canada to the other G7 countries and the OECD average for the 5 MB and 20 MB prices.


All prices are in U.S. dollars converted to Purchasing Power Parity.

Canada has the second cheapest price for the 5 MB usage and the least expensive for those using 20 MB over multiple sessions among the G7 countries.  

Compared to all the OECD countries, Canada ranked seventh in terms of lowest price for the 20 MB plan. Canada’s prices came in 13th place for the 5 MB plan, but in a virtual tie with the five countries that placed ahead of Canada.  

Canada’s mobile data roaming prices are not as far out of line as the OECD press release would have us believe. At least, not when doing the comparisons based on more realistic scenarios.

Friday
Oct082010

Internet in Canada - Adoption vs. Availability

Success in broadband internet is often gauged by the proportion of the population that is connected. 

In 2009, 80% of Canadians were going online, 77% using a connection at home. 70% of Canadians surveyed reported having a high speed connection at home.  These findings are from Statistics Canada’s Canadian Internet Use Survey, which collected data from 23,000 Canadians.

The results of the survey were released in May and included a number of interesting results. A paper presented by Statistics Canada staff at a recent conference included some further analysis. 

Among the more interesting illustrations are two that show a dramatic increase from 2007 to 2009 in the rate of adoption of high speed internet by economic regions.

 

In just two years, several additional economic regions had at least 70% of the population using a high speed internet connection from home.

A complement to the survey data is found in the information published by the CRTC as part of its annual Communications Monitoring Report

The CRTC relies mainly on data reported by internet service providers to measure the number of residential internet subscriptions.  This information can be compared to total household counts to gauge penetration rates by household, rather than population. It also has more detailed information on the type or speed of internet connection.

The following chart shows the distribution of internet connections among Canadian households.

Source: CRTC CMR 2010, Table 5.3.3 and page 137.Source: CRTC CMR 2010, Table 5.3.3 and page 137.

This information generally corroborates the Statistics Canada population surveys, indicating about 76% of households with a connection, compared to 77% reported by Statistics Canada.  The CRTC subscriber data also indicates 72% with a high-speed connection, slightly higher than the 70% of Canadians in the Statistics Canada survey.

The CRTC subscriber data also indicates that 62% of Canadian households have a broadband connection (defined as at least 1.5 Mbps downstream). More striking is the fact that 45% of all homes, or nearly one-half, have a broadband connection of 5 Mbps or greater.

The adoption of broadband internet among the population may be a popular metric for gauging success. However, the availability of broadband internet has been the focus of most government initiatives, both in Canada and elsewhere.

The reality is that adoption continues to lag availability by a considerable margin. According to the CRTC data, broadband internet is available to 95% of households in Canada, excluding services available from satellite.

Nationally, 38% of households do not subscribe to broadband internet, which is more than seven times the proportion of households that lack access to such service.

In terms of rural communities, broadband is available to 84% of households, compared to 100% in urban.  However, the gap between availability and adoption is not that different between the two regions. Data on urban versus rural adoption is for high-speed connections rather than broadband but still suggests a gap of more than 30 percentage points in either case.

Provincial breakdowns between broadband availability and adoption indicate a similar pattern, with the gap ranging from a low of 17 percentage points in the North (which also has the lowest level of availability) to a high of 42 percentage points in New Brunswick, where broadband is reported to be available to all households. 

Source: CRTC CMR 2010, Figure 5.3.7

The availability statistics are based on primarily wireline broadband technologies. Recent announcements related to mobile and satellite technologies have the potential to make broadband connections availability to all households. 

However, the Canadian experience to date indicates that other steps are needed to close the gap between adoption and availability. More research on the barriers to adoption that are not related to availability would be welcome.