| The CWA is blatantly dishonest here They get their international average download speeds, according to their report, from this ITIF study:
»www.itif.org/files/BroadbandRankings.pdf
That's the one that gives the averages of 61 Mbps in Japan, 45 Mbps in South Korea, 18 Mbps in Sweden, 17 Mbps in France and 7 Mbps in Canada. But that same report gives a value for the USA, of 4.8 Mbps, not 1.9 Mbps. (See Table 1, on page 4.)
What's the difference? Well, the ITIF report is based on advertised speeds from major vendors, geographically averaged. From page three of the report, they say:
However, measuring speed is not as straightforward as penetration because national networks are normally composed of connections of widely varying speeds. We calculate average download speeds based on OECD data that compiles the advertised speeds offered by several major broadband providers in each country. The OECD gathered this data from national providers wherever possible. This is obviously a source of bias. Not everyone who has broadband has the highest advertised speed. Plenty of people have 768k DSL even when 5 Mbps cable is advertised in their area, for example.
The CWA numbers have their own set of bias (people with slow connections might be more willing to take the Speed Matters test), but might be more accurate. However, it's completely unjustified to compare speed test numbers for the US to advertised speeds everywhere else. Particularly when the source of the international numbers gives a US number itself, more double that of the CWA number. Speed test derived numbers for other countries might indeed be slower-- as international speed test results do seem to show. |
