Some numbers one sees in reports on broadband adoption just do not make sense.
More often than you’d think, you see a headline that X percentage of people, or X millions of people, “lack internet access.” Sometimes that is accurate. Many millions of people live where fixed network internet access networks are not available.
But it is quite another thing when it is claimed that people “lack internet access” because they have chosen not to buy a particular form of internet access, when it is available. Ignore for the moment use of mobile internet access, and look only at fixed network access.
It is something else again to claim that the digital divide is as wide as “five to seven times” less “likelihood to lack adequate access” in a city with at least 95 percent fixed network internet access purchase rates, with the lowest district rates being 93 percent take rates.
So here’s a headline I recently ran across: “Low-income areas five times as likely to lack internet access.” Here is the study that supports the claim.
My problem is that I am having a hard time figuring out, in a city where the lowest reported take rates are 93 percent of homes, where the “five times” to “seven times” gap comes from.
The report states that “95 percent of households report internet access in the place where they live (an increase of 10% since 2014).” That ranges from a high of 97 percent to a low of 93 percent in the city.
It might be correct, at a national and high level, that living in poverty is a risk factor for households not buying fixed internet access, as the study indicates. But the study’s own data shows that 75 percent of Seattle homes defined as “living in poverty” do buy internet access.
It arguably is true that buy rates tend to be lower for homes with a member of the household living with a disability, or homes where English is not the primary language, or households headed by someone older than 65, households containing only single people or minority households.
But one sees relatively little of differences of that magnitude in Seattle, despite the report’s claims.
“However, we are also seeing significant gaps in access, particularly in low-income and insecurely-housed populations. People living in these communities are five to seven times more likely to lack adequate access to the internet than the average Seattle resident.”
The report’s data seem not to suggest actual behavior of that level. Ignore for the moment that “adequate access” is in other reports a measure of lack of physical facilities. What we are looking at here is not “access,” but “buy rates.
According to the study, households with a disabled member have 85 percent buy rates. Non-English households have 90 percent buy rates. Households with an older adult have 91 percent buy rates, as do households with single adults living in them. Households of racial or ethnic minorities have 92 percent buy rates.
Yes, there are gaps in buy rates. But it is hard to make the math work for a claim of “five to seven times” greater likelihood to lack adequate access.
Methodologically, the study also assumes that the cost of internet access is the price paid for any service, including an internet access service, or a bundle including other services. That is a novel way of describing internet access cost, akin to the adage of “comparing apples to oranges.”
The study claims “households pay on average $150 per month for internet service.” Uh, not really. The study itself says that figure includes money spent by households to “access the internet and internet-related services.”
That means, if the household buys a bundled service, the $150 includes internet access, entertainment video and possibly fixed network voice, in some cases. Odd is the inclusion of video and voice as part of the fee to use the internet.
No comments:
Post a Comment