Measuring Broadband Performance
Here’s the video from yesterday’s release event for the FCC’s new study on broadband speeds in the US.
It’s not every day that a government agency releases a study that corrects its previous work, but that’s what happened here and the policy discourse is all the better for it.
Previous Post
Not a government sponsored study, but Ookla has 6 million unique visitors a month from people wanting to test their broadband speeds worldwide. They rank the United States far higher than Europe, Asia Pacific, G8, and the OECD. http://www.netindex.com/promise/
The Ookla study has the same problem that the comScore study does: they don’t know what speed the consumer is supposed to be getting, don’t correct for other activities, and use a self-selected sample. That’s not the way to measure BB performance.
Ookla asks the customer for the price and advertised speed of their broadband service. This is how they got their data on value of service and honesty data.
The issue of being self selected is an issue if the sample size was small, but having so many samples allow the analyst to pull a statistically valid subset.
This is why the Ookla data is very similar to the FCC study.
I see Japan and South Korea are not included in Ookla’s results due to “insufficient data.” (i.e., broadband in those countries is so fast that users don’t feel a need to test) I’d expect they’d whup everyone if they were included.
The FCC found that most people don’t know what service tier they subscribe to, so you really can’t just ask people, you have to verify; see: John Horrigan and Ellen Satterwhite, Americans’ Perspectives on Online Connection
Speeds for Home and Mobile Devices, 1 (FCC 2010), at http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-298516A1.doc (finding that eighty percent of broadband consumers did not know what speed they had
purchased).
The other objections still apply, and the conclusion is that the Ookla reports fall far short of the rigor of FCC’s work.