Assessing Early Public Response to COVID-19-Related Restrictions in New York City Using Spatial Analysis of Urban Mobility Data

The rapid spread of COVID-19 in the United States initiated shelter-in-place policies that significantly impacted human mobility and daily routines starting in March 2020. Prior literature has examined the differences in lockdown policy efficacy and compliance with government orders1-6, as well as the effect of mobility changes on case counts7-12. However, less attention has been placed on the connection between mobility and socio-demographics after the onset of COVID-19 within a city’s borders. This paper focused on how human mobility patterns in New York City during the first three months of the pandemic differed based on socio-demographic factors like age, household income, and method of transportation to work. A secondary analysis determined if the four measurements of mobility used, namely distance traveled from home, home dwell time, non-home dwell time, and percentage time home, yielded significantly different findings. A mobility ratio representing the change in mobility between the first two weeks of February and April 2020 was created using aggregated and anonymized cellphone mobility data from SafeGraph. A Global Moran’s Index was calculated for each mobility ratio to test for the presence of spatial autocorrelation, and then two spatial lag models were applied to account for the existence of autocorrelation. That there existed significant differences in mobility patterns based on socio-demographics reinforced the need for physical distancing policies that acknowledge the demographic diversity present not only between but also within cities.