Staying at home is a privilege: Evidence from mobile phone location data in the U.S. during the pandemic

July 1, 2021

When lockdowns began in the spring of 2020 few people regarded the ability to comply with them as a sign of privilege. However, that’s what researchers Xiao Huang (Dept. of Geoscience, University of Arkansas), Junyu Lu (School of Community Resources and Development, Arizona State University), Song Gao (Dept. of Geography, Univ. of Wisconsin), Sicheng Wang (Univ. of South Carolina and Edward J. Bloustein School of Planning and Public Policy, Rutgers University), Zhewei Liu (Dept. of Land Surveying and Geo-Informatics, Hong Kong Polytechnic University), and Hanxue Wei (Dept. of City and Regional Planning, Cornell University) discovered in their research “Staying at Home Is a Privilege: Evidence from Fine-grained Mobile Phone Location Data in the United States during the COVID-19 Pandemic,” ( Annals of the American Association of Geographers,

An MSA (metropolitan statistical area) is delineated as a region that consists of at least one urbanized area with a minimum population of 50,000 (U.S. Census Bureau 2020). The area defined by the MSA generally received similar mitigation measures during the COVID-19 pandemic and is typically marked by significant social and economic interaction, thus serving as an ideal geographic unit in this study. In addition, the dense population in MSAs ensures sufficient home-dwelling records. The geographical boundaries of these MSAs are the 2019 TIGER/Line Shapefile products issued by the U.S. Census Bureau.

Data from millions of mobile phones at the CBG (Census Block Group) level were collected in the 12 largest, most populated MSAs.  The researchers found statistically significant correlations between the increase in home-dwelling time and variables that describe economic status in all MSAs.  There was a positive association between income and the ability to stay home.  Poor communities tend to show less compliance with stay-at-home orders because they could not afford to absorb the economic shocks of the pandemic. 

The percentage of school children (ages five to seventeen) also plays an important role in the ability to stay at home.  Those without daycare and the inability to work from home struggled with childcare. Another correlation occurred around educational attainment. Households with high educational attainment were able to be flexible and adapt to working from home while those households with low educational attainment did not have that option and frequently had to choose between working and paying for food and rent.

Additionally, disparate compliance with stay-at-home orders leads to disparities in exposure to COVID-19.  Disparate exposure for vulnerable populations can further compound other disadvantages, such as underlying comorbidities, poor access to low and low utilization of high-quality health care, and limited access to COVID-19 testing centers; this leads to negative health outcomes for vulnerable populations.

The study points out that stay-at-home orders are more easily complied with if one is educated, well-off, and white.  Those who have less education, are poorer, and of color face loss of income and/or dangerous exposure to COVID-19 and have poorer health outcomes when infected with COVID-19 due to healthcare inequities both before and during the pandemic.

It is important to ensure the equitable allocation of health care resources and financial resources, such as subsidies for more vulnerable populations. The systemic social inequity issues call for a high-priority assessment of the long-term impact of COVID-10 on socially disadvantaged groups.

Recent Posts

NJSPL – New Jersey Employment Concerns Revisited

As 2024 began with yet another surprisingly strong jobs report for the U.S. (353,000 jobs added in January and the unemployment rate steady at 3.7%), and with a full year’s worth of 2023 state-level employment data now available, it’s worth briefly revisiting some of...

New Research: The Traffic Calming Effect of Delineated Bicycle Lanes

Abstract We analyze the effect of a bicycle lane on traffic speeds. Computer vision techniques are used to detect and classify the speed and trajectory of over 9,000 motor-vehicles at an intersection that was part of a pilot demonstration in which a bicycle lane was...

Upcoming Events

Event Series CAREERS

Career Virtual Drop-ins


Bloustein Career Development Specialists Cheryl Egan and Andrea Garrido will be in a Zoom Room on Monday's beginning January 22, 2024 (excluding holidays and spring break) to answer questions, provide […]