Connecting the Dots: Using Big Data to Understand Mobility Gaps in American Cities

Socioeconomic inequality is a key issue in cities across the globe. While many studies focus on income, education, healthcare, and housing affordability, research on the so-called mobility gap—how different socioeconomic groups move about the same city—has been limited in scope. One reason for this is that traditional datasets (e.g., the Census, built environment data, economic data, health outcomes) are relatively static: they are too small and evolve too slowly to properly capture mobility trends or allow policymakers to address inequalities in real-time.

A study by the Bloomberg Harvard City Leadership Initiative used smartphone location data to examine the mobility patterns of nearly 13 million individuals living in nine cities across the United States before and during COVID-19. Unlike other recent studies using smartphone GPS data, which have focused on one destination at a time, we were able to paint a more complex picture of urban travel by considering several destinations at once. This approach allowed us to better understand how mobility trends vary depending on residents’ socioeconomic status (which we measured through a multi-dimensional socioeconomic index, or “SEI”), and how different factors—such as walkability of neighborhoods, proximity to public transportation, and the disruptive effect of the COVID-19 pandemic—can impact mobility gaps.

The study is an example of an emerging field driven by newly accessible, high-volume, high-velocity data that can provide a much more granular, timely understanding of human behavior than traditional datasets. For city leaders interested in improving the quality of life for their residents, this kind of analysis can help shine a light on the blind spots that may arise when we look only at built environments and traditional social policies.

Mobility Gaps in American Cities before and during COVID-19

We examined how mobility differed between the lowest- and highest-SEI groups in Atlanta, Boston, Chicago, Denver, Houston, Los Angeles, Philadelphia, Seattle, and St. Louis in 2019 and 2020. Specifically, we studied differences in mean trip distances, the proportion of trips made to three types of destinations (parks, healthcare, and retail), and whether home location—walkability and proximity to public transit—affected mobility outcomes. We used data provided by SafeGraph, a company that captured the mobility patterns of smartphone users and made this information freely available for public interest purposes during the pandemic. (SafeGraph no longer provides mobility data, but other companies offer it for a fee.)

We found striking differences in the mobility trends of the least- and most-advantaged residents. Among other things, low-SEI residents made fewer trips to parks and more retail trips in 2019, and these differences grew in 2020. While disadvantaged residents made more trips to healthcare facilities before COVID-19, this difference flipped in 2020, with low-SEI groups reducing their healthcare visits while high-SEI groups increased them.

Home location affected each group differently. Neighborhood walkability allowed better-off city dwellers to make fewer and shorter trips during COVID-19 but conferred no such benefit on low-SEI residents. High-SEI residents were more likely to travel by car even when public transit was easily accessible.

Mobility patterns also differed across cities. While the pandemic widened mobility gaps significantly in seven of the nine cities studied, Seattle and Atlanta experienced much smaller pandemic-induced changes in their mobility gaps. This disparity raises interesting questions about how the particular spatial, social, and economic aspects of cities affect mobility.

Takeaways

Big data, like the cellphone location data highlighted in our study, offer city leaders insights into closing mobility gaps that can inform policies beyond transportation, such as investing in access to childcare, schools, and medical facilities; creating equitable opportunities to use amenities, services, and public spaces across neighborhoods; and extending affordable housing programs to commercial spaces so that low-SEI residents living in walkable neighborhoods can afford nearby goods and services.

More broadly, however, big data can help city leaders better understand their particular city’s mobility patterns to help devise solutions that can best address the specific challenges they face. To boost your city’s data-analytic capacity, consider the following: