While broadband access has become increasingly critical in the twenty-first century, the COVID-19 pandemic’s stay-at-home mandates accelerated the necessity of such access. In light of this development, our current work explores the question of, "does a region’s differential access to and quality of broadband influence the short-term impacts on employment during the COVID-19 pandemic?"

We use econometric difference-in-difference methods at a county-level spatial analysis paired with robust broadband measures that reach beyond the standard advertised speeds as provided by the Federal Communications Commission, including measures provided by Microsoft, ACS and FCC. Our initial findings show that when all else is equal, after COVID, counties with more than 50% population access to 25 mbps download and 3 mbps upload experience an increase of 1.23% in their unemployment rate over similar counties that have less than 50% access to broadband. We are now in the midst of exploring the potential mechanisms which could explain the reason for this. We think there is a strong opportunity for an undergraduate student to take ownership of several pieces of the current exploration, creating space for their own unique analysis within this work. These pieces include:

  1. Researching and understanding the Ookla and MLab datasets, which are relatively new and evolving datasets in this space, and provide a measured assessment of broadband access throughout the United States.
  2. Learning how to use the Ookla and MLab APIs in order to collect the most up-to-date data from these data sources. This would include pulling access at the county and census-tract level for 25 mbps/3 mbps and 100 mbps/10 mbps where possible.
  3. Learning how to use the census API in order to ingest census-tract data for demographic purposes.
  4. Learning the basic assumptions of difference-in-difference methods so as to be able to run the existing models using the newly ingested Ookla, MLab and census tract demographic data. We would provide the remaining additional data required to run this work.

This approach would allow the student to take ownership over a specific, tangible portion of the project. While expected to be self-directing, the undergraduate student will work directly with a Ph.D. student, Nikki Ritsch, on a weekly basis, meeting as often as is needed for the work required. The student will also meet bi-weekly with the full research group in addition to attending individual meetings with the PI as needed. 

This work contributes to a new research area of rising importance on the impact that infrastructure has on social systems amidst extreme events, such as pandemics, and the broader implications therein. As such, it can help ascertain where to begin broadband expansion goals and may shed light on where the Emergency Broadband Benefit could be most effectively implemented in order to reduce inequity in broadband access. In this manner, this study treats broadband infrastructure as a “social sensor” for understanding where a lack of access amidst COVID-19 is most acute and, therefore, where to first target policy resources.

For additional information, students may also contact Nikki Ritsch. Ritsch is a Ph.D. student with Daniel Armanios, and students who are accepted for this opportunity would be working directly with her as well as Armanios. nritsch@andrew.cmu.edu.