When "Data" Does Not Match "Perceived Reality" Perhaps the Data or Your Perceptions are "Wrong"

Trying to figure out what people really pay for home broadband is tricky. Prices differ by provider, location and the number of competitors in a market. Prices also vary by the level of government subsidies and take rates of such subsidies. 

One has to decide which plans to compare, and those choices shape the outcomes. The other issue is the inability to adjust the analysis for discounts and promotions that affect what customers actually pay. In other words, the retail tariffs we choose to compare  are not necessarily reflective of active consumer behavior. 

The most recent example of this is an analysis of home broadband costs by the Internatiomal Telecommunications Union, which reports that the “lowest-priced home broadband plan” offered by the largest U.S. supplier average more than $130 a month.

If you follow U.S. broadband prices, you know that is incorrect. Since the largest U.S. ISP is Comcast, the ITU must have looked at Comcast’s stated prices. And if you check, you can see that the stated retail price (after a promotional period) does jibe with the ITU figures. 

source: ITU 

 

The issue is that Comcast customers do not seem to be paying triple-digits for the “lowest-cost” home broadband plans. 

That seems wildly incorrect. Methodology is an issue. The lowest-cost budget plans typically are in the $10 to $15 a month range and support speeds around 50 Mbps (moving higher, as do speeds on all plans). 

Those are the plans ITU should have reviewed were it looking at the comparable “lowest-cost plans offering 5 Gbytes of usage and minimum speeds.” 

source: HighSpeedInternet.com. 

But those plans are rarely listed on websites showing available plans. You would have to hunt to find plans offered by all leading ISPs for low-income households. 

Instead, the ITU researchers seemingly looked at prices shown on the Comcast websites that do not represent the comparable lowest-cost plans. To be sure, Comcast, the largest U.S. ISP, shows charges of $30 a month for services operating at 100 Mbps. 

And, to be sure, Comcast also says those prices are good only for one year, with sharp price increases after 12 months. Comcast says its 100-Mbps plan will grow to $81 a month after 12 months. 

The issue is that it would be hard to find anybody who actually pays that amount for a 100-Mbps service, even after a 12-month period. 

The average U.S. home broadband service  costs about $64 a month. If the cost of the lowest-priced plan really were more than $100 a month, as the ITU analysis suggests, the “average” U.S. price could not be as low as $64. By definition, the average would have to be much higher. 

According to Openvault, only about 20 percent of U.S. households purchased services operating at 100 Mbps or less in the second quarter of 2021 and only 18 percent in the third quarter of 2021 and 17 percent by the fourth quarter of 2021. 

source: Openvault 

And that is why methodology is so important. Actual measurements of home broadband speed show only 17 percent of subscribers are provisioned for speeds less than 100 Mbps. Only nine percent are provisioned for speeds of 50 Mbps or slower. 

The point is that any analysis of home broadband focused on the lowest tier of service, in terms of speed or price, would not tell you very much, in any country. What is arguably much more useful is an analysis of the “typical” plans customers buy, and not the highest or lowest price plans; fastest or slowest speed options available. 

Unless one is clear about methodology, it is easy to make unclear or misleading statements, such as “X percent of customers do not have access to Y speeds.” Does that mean such customers cannot buy because the service is not available? Or does it mean they could buy, because the service is available, but they choose not to buy, preferring some other plan?

The answers matter. 

I may choose not to buy a Tesla. That does not mean I cannot buy a Tesla. Some will rightly argue that home broadband is a necessity, and a Tesla is not. Noted. But virtually all the global data shows that, over time, the cost of home broadband globally has declined, measured as a percentage of gross national income per person. 

To be sure, according to the ITU, fixed broadband prices (adjusted using the purchasing power parity method) have risen since 2015, after dropping since 2008, while mobile data costs have dropped steadily. 

But 2020 prices were still lower than in 2008, and that assumes we accept the data as accurate, which I do not. If the same methodological issue applies to other markets, then prices are overstated. 

Beyond all that, there are hedonic adjustments, referring to the change in a product’s performance over time. Beyond price, the performance of our smartphones, personal computers or home broadband are vastly different than they were 20 years ago. 

Is the $300-per month 756 kbps internet access connection I was buying about 1996 the same product as the gigabit connection I now buy that costs possibly $85 a month? Is that gigabit connection the same product as the 300-Mbps connection I was buying a year ago, even if that product “cost less” than the gigabit connection?

Methodology always matters when evaluating home broadband availability, quality and cost. In this case, the ITU analysis seems quite flawed.