There was an interesting paper published in May – Who is  using the internet at faster speeds? The authors are Roberto Gallardo, the vice President for Engagement at Purdue University and Brian Whitacre, a Professor and the Jean & Patsy Neustadt Chair in the department of Agricultural Economics at Oklahoma State University.

The paper highlights some interesting observations that are contrary to conventional wisdom about broadband. The studies showed that the higher the percentage of Black non-Hispanics and Hispanics in a community, the higher the average download broadband speed. Conversely, the higher the percentage of White non-Hispanics, the lower the average broadband speeds.

The team had originally done a study based on 2021 speed tests that had shown this result. They were surprised by the results and wondered if the pandemic had some impact on 2021 data that might have skewed the findings. In a second study they looked at 2019 speed tests from before the pandemic as well as 2022 speed test data which is past the worst periods of the pandemic.

The second study confirmed the original findings. When the percentage of Black non-Hispanics and Hispanics increased by one percentage point, the average download speeds increased by roughly 5 Mbps in 2019 and rose to almost 7 Mbps and 12 Mbps in 2022, respectively.

The study also correlated factors like race and the presence of children in a home. There was a significant increase across the board in broadband speeds between 2019 and 2022 for all households with children. The study found that the impact was greater for the three years for Black non-Hispanics (almost 18 Mbps faster) compared to White non-Hispanics and Hispanics (roughly 14 Mbps faster for both groups).

Their overall conclusions were as follows:

• Rural, older, and poorer groups are on the wrong side of the digital divide.
• Having children in the home correlates with having faster broadband speeds than neighboring homes with no children.
• There was a big difference before and after the pandemic in the broadband speeds in homes where people worked at home.
• Higher education correlated with faster Internet use, particularly for upload speeds.
• White non-Hispanics seem to be increasingly on the wrong side of the digital divide.

The team acknowledged that understanding the broadband landscape is complex and in all populations there were 20% to 30% of speed tests that could not be explained by the findings. The analysis also had no way to distinguish between households that were able to choose their broadband speed from multiple ISP options versus homes where there was a limited number of ISP options.

Like most academic papers, the team did not speculated on the reasons behind their findings, and just reported the statistics they uncovered. My firm has been doing broadband surveys and detailed interviews with residents for many years, and some of these results were not a surprise to me.

For example, I’ve heard many stories from homes with children who sacrifice to buy faster broadband speeds for their children. I was also not surprised to hear that those working from home have faster broadband speeds – because that is a prerequisite for landing such jobs. I’ve run across many stories of folks who have moved to where broadband is faster in order to work from home (or who moved away from areas with poor broadband).

It’s harder to understand why some of the other findings are true, such as the one that showed that the greater the percentage of Black non-Hispanics and Hispanics, the faster the broadband speeds. I’ve encountered counties with a big difference in the overall desire to have good broadband. For example, I’ve studied a few rural counties where a lot of residents took pride in not connecting to the Internet, while other counties in the same state had similar residents that were begging to get faster rural broadband. I haven’t the slightest idea how to correlate such widely different behavior with the national trends spotted by Gallardo and Whitacre, but I hope they keep digging to help folks like me understand the incredibly complex broadband landscape.