Stark digital divide in US internet speed amid Covid-19 crisis

A study by global edge cloud platform Fastly looking at internet performance within median income brackets in the US has found a clear stratification of download speeds, indicating households with lower incomes are experiencing lower download speeds compared with households on higher incomes.

The study, carried out from February to April 2020, is the first in a series of data analyses on what the company sees as a digital divide – the unequal distribution of internet access across socio-economic and geographic lines.

To understand the impact of the digital divide on low-income families during the Covid-19 coronavirus pandemic, Fastly used as a benchmark the 5-25Mbps range of bandwidth recommended by US regulator the Federal Communications Commission (FCC) to support students and telecommuting. It compared the median download speed between 26 February and 21 April for five brackets of annual average income.

And while the company hypothesised that internet performance would indeed vary by income, it admitted to being surprised to find what it called a stark stratification in user experience.

“Because of Fastly’s visibility into telecommunications and ISP [internet service provider] activity, as well as end user experience, we can connect threads that help to define the tangibility of the digital divide for many communities in the US,” said Fastly’s chief architect and founder, Artur Bergman.

“This becomes especially insightful as more and more people transition to a more web-reliant world in light of Covid-19,” he added. “We observed relief for those segments of the population arise as a result of decisions from ISPs like Comcast and Cox Communications. Both implemented upgraded internet speeds in mid-March, which led to recognisable improvements in download speed for lower-income communities.”

The analysis showed locations with median incomes in the highest brackets are seeing the highest download speeds, compared with those on the lowest median incomes, which are seeing the lowest download speeds. This reflects that, among other inequities, people in higher income neighbourhoods have both access to higher-bandwidth connectivity and the means to use it.

During the date range, Fastly saw an increase in download speed in users of all income groups, with users in the lowest income group seeing the largest increase. This bracket had its median download speed increased from 13.7Mbps to 17Mbps, a 25% increase in download speed over a period of two months.

Yet the analysis also showed that on 26 February, 26.3% of the connections in the lower income bracket did not meet the 5Mbps minimum download speed recommended by the FCC. However, from mid-March to mid-April, Fastly observed the biggest improvement in number of users without the minimum required download speeds in that same bracket, dropping from 26.3% to 22.3%.

The US ISPs’ launches had clearly observable effects. The 14 March decision by Comcast to increase the speed of its Essentials bundle, offered to all qualified low-income households for less than $10 a month, from 15Mbps to 25Mbps. Prior to the announcement, the median user in the lowest income bracket saw download speeds at 27.2% lower than the median user in the highest income bracket. Following the announcement, the download speed gap between a median user in the lowest compared to the highest income bracket reduced from 27.2% to 13%.

On the same day, Cox Communications upgraded its residential internet packages to 50Mbps for 60 days. Prior to the announcement, the lowest income median user saw download speeds at 13.9% lower than the median user in the highest income bracket. After the announcement, the download speed gap between a median user in the lowest compared to the highest income bracket reduced to 1.8%.

In what it suggested was a likely result of these ISP decisions, Fastly calculated that the median download speed of all connections in the US increased from 17.9Mbps on 26 February to 19Mbps on 21 April – a 6% increase over a period of two months.