The influence of higher ed testing on Hampshire County COVID-19 data

This is a “white paper” we wrote to show that the number of tests in the 5 Colleges is 90% of the number of tests in Hampshire County, and that this might be having a dramatic effect on positivity rates. It was covered in a Jan. 23 Hampshire Gazette article. We are circulating it to encourage discussion of how to improve local data reporting amongst local and state officials, and amongst the broader community. A list of officials we have shared it with is appended at the bottom of this webpage. We encourage others to share it as well. If you share it with other officials, please write to so that it can be documented here.

Pater, Joe, Michael Stein, and Susan Voss. 2021. How might higher-ed COVID-19 asymptomatic testing influence testing rates and percent positivity in Hampshire County? Ms. January 17 2021, University of Massachusetts Amherst and Smith College.

Executive Summary

This white paper explores the impact of higher education testing on the COVID-19 data reported by the state for Hampshire county. We show that in Hampshire County, home to the Five College Consortium, the data are dominated by higher-ed testing. The number of tests associated with testing at the Five Colleges is about 90% of the number of total tests in Hampshire County. This means that the test numbers provided by the state are likely not adequately representing the broader community, and may be concealing inadequate testing capacity for community members outside higher-ed.

Because higher-ed testing involves repeated testing of asymptomatic people, it has much lower percent positivity rates than testing of the general public. Its dominance in the Hampshire County data appears to skew positivity rates. Specifically, for the period between 8/26/20 – 1/14/21, we find that the percent positivity rate for Hampshire County including higher education testing is 1.17%, but with higher education testing removed jumps to about 4.56%.

We also demonstrate how the effect of higher-ed testing can impact a municipality within the county by exploring the impact of Smith College testing on the percent positivity rate of Northampton, MA. We find that with Smith testing removed, Northampton’s percent positivity rate rises from 3.36 to 4.75 in the week ending 1/14.

Because there are currently no publicly available data on the towns or cities of residence associated with the Five College tests, our estimates of the effect of higher-ed testing are preliminary. Nonetheless, it seems likely from these estimates that the dominance of higher-ed testing is impacting the classification of towns and cities in terms of the state’s color-coded risk metric.

We urge state and local leaders to:

  • Determine if Hampshire county residents who are not employed within a higher-ed institution have adequate access to COVID-19 testing.
  • Publicly share appropriate local data so that citizens and local leaders understand the uncertainty in the state-reported data and are able to make local data-driven decisions.
  • Offer guidance to counties and municipalities with significant higher-ed testing on how to interpret their local percent positivity values in regard to public health guidance.

Note: the Jan. 1, 2021 version that was circulated before the new version of Jan. 17th included an error in data that was provided to us by the City of Northampton. 

Shared Jan. 1 with Northampton officials: Mayor Narkewicz, NPS Superintendent Provost, Senator Comerford, Representative Sabadosa, Health Director O’Leary, and Board of Health Chair Levin.

Shared Jan. 2 with all Northampton City Council members.

Shared Jan. 4 with all NPS School Committee members.

Update Feb. 16 We obtained summary testing data for all the higher ed institutions in Massachusetts up to Feb. 8th.