Our Internet is Getting Faster…Yay?
We are constantly getting bombarded with advertisements by cellular carriers and broadband companies promising faster speeds for (usually) less money. Overall, this is good for us as consumers. Now, advertised internet speeds from the mid 2000s seem slow and unacceptable. Apps and websites are more bandwidth-demanding than ever, and it’s a good thing that ISPs at least trying to keep up with this demand.
In a recent article by The Verge, they point out that some of the logic that the White House used in a recent report on faster fixed broadband speeds was a little “fuzzy”. In the report, the White House attributes these increases in fixed broadbands speeds to the abolishment of net neutrality laws. However, The Verge points out that this is unlikely as for some of the metrics, the law was still in effect and others are more easily attributed to factors such as cheap upgrades to DOCSIS 3.1 that cable companies have been undergoing for several years and community broadband efforts.
However, this is an very nuanced issue and there are a significant amount of factors that could be contributing to these speed increases. Getting to the bottom of these causes is also quite nuanced, especially because one of the biggest databases of internet speed tests is owned by Ookla, and they don’t make the raw data publicly available, so all we have to use from their dataset is aggregated, metropolitan area data.
The FCC does, however, keep a massive database of all registered ISPs, their advertised internet speeds and the locations in which they operate. While we all know that advertised speeds typically aren’t what we actually get in our homes, it’s as close as we can get in this case.
The database only has data for 2015, 2016, and 2017, but this gives us a good idea of where fixed broadband stands relative to the past couple of years:
Surprisingly, the national average advertised upload and download speeds have almost doubled between 2016 and 2017. Immediately, I thought that this must be because one state had a large deployment of gigabit fiber or something and was skewing the average, but that doesn’t seem to be the case.
It seems that, save for a few states, advertised download speeds have increased significantly across the country. Unfortunately, it’s still tough to determine why this happened. The law that designated fixed broadband companies as common carriers was repealed at the very end of 2017, so it’s possible that these speed gains were realized because of net neutrality law, but it’s hard to prove this either way.
While national and state averages give us a good idea of how internet speeds have changed for the average citizen, they aren’t a great indication of how wide these increases are felt. Typically, those in metro areas have more options for broadband access and those options are typically faster. While the FCC data doesn’t include any metrics on whether the area that a provider services is rural or urban, it does include FIPS codes that can be joined to census data. RUCA (Rural-Urban Commuting Area) codes classify census tracts according to how much of that population commutes, giving us a good sense of how rural or urban an area is.
Because of the computing intensity of this joining process (the FCC data has over 70 million rows by itself and joining the RUCA codes produces a ~17GB dataset), I only did this analysis for the entire nation using the 2017 data.
The RUCA codes generally work like this: 1 is the heart of a metropolitan area and 10 is completely rural, with several variations in between. 99 is reserved for unclassified census tracts. As you can see, the theory that those closer to metropolitan areas tend to get faster internet speeds mostly holds up.
Colorado as a case study
Since working with all 70 million rows in the dataset, I took Colorado as a small case study.
Using the same RUCA codes as above, we can see that download speed improvements are seen across the board in Colorado, with the biggest increases in speed in the metro areas. In the article from The Verge, a probable cause for internet speed increases is new community broadband efforts, which are mostly in more rural areas where other bigger broadband companies lack coverage. In the above chart, the rural download speeds (at least in Colorado) did not see the same gains and as the metro areas. This points to a problem that continues to exist in rural areas. This seems to indicate that even if community broadband efforts are expanding in more rural areas, they aren’t exceeding the advertised speeds of previously available ISPs.
Wrap up
While the data from the FCC are not real speed tests and just advertised speeds, we were able to gather that internet speeds across the country have increased since 2015. It’s hard to draw any hard conclusions about any causes for these gains, but even with the limited amount of data, there are findings that support both The Verge’s and the White House’s claims.
