Historical Redlining and Park Use during the COVID-19 Pandemic.docx

<strong>Background: </strong>Despite the health benefits of urban green space, disparities in its access and use have long existed. Emerging evidence suggests an adverse impact of redlining, a discriminatory practice decades ago, on multiple health outcomes. However, whether and to what degree redlining contributes to these disparities remains unknown particularly during a pandemic. With newly available mobility data tracking the locations of large numbers of mobile devices, this study links historical redlining with changes in green space use during the COVID-19 pandemic. <strong>Objective: </strong>This study examines how changes in park visits during the lockdown period (3/23/2020-8/2/2020) are associated with redlining across census tracts in three large U.S. cities. <strong>Methods: </strong>HOLC neighborhood redlining grade data were merged with SafeGraph mobility data at census tract level for New York City, Chicago, and Philadelphia. Ordinary Least Square regressions were conducted to assess the association between dominant redlining grade and relative change in park visits in census tracts by comparing the lockdown period to the reference period. Spatial error and lag models were also used to account for potential spatial autocorrelation. <strong>Results: </strong>Park visits during the lockdown period in 2020 decreased by at least one-third in the three cities. The influence of redlining varied across neighborhoods and cities. In New York City, neighborhoods with more redlined areas experienced the largest drop, sharper decreases concentrated in neighborhoods previously graded as “best” or “still desirable” in Philadelphia, but the effect was barely present in Chicago. In addition, changes in park visits are positively correlated between neighborhoods in New York City and Chicago, but it’s not observed in Philadelphia. <strong>Significance: </strong>The link between historical redlining and current green space use is context-specific. As neighborhood dynamics vary across places, more decreases in park visits were observed in both historically lower-graded (New York City) and higher-graded neighborhoods (Philadelphia).