We have been living in strange times.
With most of the economy shutdown since mid-March, families throughout the United States have adapted to their new circumstances.
As a public health policy researcher, the ongoing pandemic has kept me at home and thinking about the same sort of questions that everyone has been wondering about over the past weeks – How prevalent is it? How deadly is it? Are we responding appropriately? How long will the threat last?
The shutdown of non-essential businesses and the issuance of a “stay-at-home” order represent unprecedented acts on the part of government to protect public health. However, despite these strong measures, certain pockets of our economy have continued to operate without much additional regulation – most notably grocery stores, supercenters and home improvement stores.
Although the stay-at-home order limited customer capacity in these stores, it did not mandate the use of masks (or other face coverings) by customers. Anyone who has been to one of these stores over the past few weeks can tell you that mask wearing has been far from universal under the current guidance.
Last week I decided it was time to explore this issue further, so I approached my 17-year old son and said “Hey, want to write a paper with me?” He asked, “what’s it on?” and I said “Masks, who’s wearing them and who isn’t.”
He said something like “Sure, what are we going to do, sit in the parking lot and watch people?” and I replied “that’s exactly what we’re doing.”
So in the space of a week, we conceived of a research question, collected and analyzed data, and are now disseminating our findings.
Between April 22 and 25, we collected data on 500 individuals at a grocery store and superstore in North Shelby County in the Birmingham metro area. As customers exited the store, we recorded their mask status, gender, age group, race/ethnicity and use of a shopping cart. We denoted individuals that were shopping together as we hypothesized that customers shopping in a group may be less likely to wear masks.
The use of a shopping cart provided a reasonable proxy for the amount of time spent in the store.
Overall, 44.2% of customers were observed wearing masks as they exited the two stores. We used a simple multivariable linear regression model to estimate the association between individual characteristics and mask use.
Women were 7.6 percentage points more likely than men to wear to wear masks. We did not observe statistically significant differences in mask use by race. Relative to younger middle-aged (appearing 30-49) customers, elderly shoppers (appearing 65+) were 40.6 percentage points more likely to wear masks, while younger adults (appearing aged 20-29) were 26.5 percentage points less likely to wear masks.
Customers who used carts were 9.9% percentage points more likely to use masks relative to customers that did not. Customers who shopped with at least one other person were 14.8 percentage points less likely to wear masks. Finally, customers at the superstore were 13.4 percentage points less likely to wear masks relative to customers at the traditional grocery store.
There are several limitations of our work, the first two of which relate to potential measurement error in our data collection. First, we are making subjective assessments of gender, race and age from a parked car at a distance of 50-75 feet. Despite both of us having 20:20 vision, it’s likely we made some mistakes here and there.
We think the measurement error is very limited for gender, fairly limited for race/ethnicity and modest for age category. The second potential limitation is that we observe people as they exit the store.
If individuals remove their face covering while still in the store we might be underestimating overall use.
The final limitation of our study is its generalizability, as both the overall prevalence of mask usage and the relationship between individual characteristics and mask usage may be unique to the two stores we studied.
Putting these limitations aside, our study demonstrates the limited effectiveness of voluntary measures to slow the spread of COVID-19.
This is not surprising given the inconsistent messages that have been provided to the public regarding face coverings over the past two months. Concerns of mask shortages for health care workers initially led the CDC to downplay the value of face coverings. On April 3, the CDC reversed course and encouraged the use of homemade masks, while continuing to discourage the use of more effective surgical-grade masks. This imperfect roll out of information has likely led to uncertainty regarding the benefits of homemade face coverings.
We are not experts on aerosol transmission of COVID, or the effectiveness of these alternative face coverings. Our data simply speak to how people have behaved in the face of this uncertainty where mask wearing is voluntary. Our work suggests that customers are making decisions based on their own perceived risk with older shoppers and individuals who spend more time in stores being more likely to wear masks.
Mayor Woodfin recently announced that face coverings would be mandatory in all public places within the city of Birmingham effective May 1st. As we take the first steps to reopen the economy, this seems a worthy sacrifice to enable a safer and swifter return of the missing pieces of our lives.
David Becker is a health economist and professor in the UAB School of Public Health; James Becker is a junior at Oak Mountain High School; All views expressed in this article are their own.
Updated May 2, 2020, to correct misspelled name in byline.
