The characteristics of multi-source mobility datasets and how they reveal the luxury nature of social distancing in the U.S. during the COVID-19 pandemic

Abstract

This study reveals the human mobility from various sources and the luxury nature of social distancing in the U.S during the COVID-19 pandemic by highlighting the disparities in mobility dynamics from lower-income and upper-income counties. We collect, process, and compute mobility data from four different sources. We further design a Responsive Index ($RI$) based on the time series of mobility change percentages to quantify the general degree of mobility-based responsiveness to COVID-19 at the U.S. county level. We find statistically significant positive correlations in the $RI$ between either two data sources, revealing their general similarity, albeit with varying Pearson’s $r$ coefficients. Despite the similarity, however, mobility from each source presents unique and even contrasting characteristics, in part demonstrating the multifaceted nature of human mobility. The results suggest that counties with higher income tend to react more aggressively in terms of reducing more mobility in response to the COVID-19 pandemic. Most states present a positive difference in $RI{\rm \; }$between their upper-income and lower-income counties, where diverging patterns in time series of mobility changes percentages can be found. The findings shed light on not only the characteristics of multi-source mobility data but also the mobility patterns in tandem with the economic disparity.

Keywords

• covid-19
• mobility
• responsive index
• visualization
• big data