ABSTRACT
During the spring and summer of 2020, cities across the world responded to the global COVID-19 pandemic by converting roadway facili-ties into open pedestrian spaces. These conversions improved access to public open space, but measuring the variation in that improvement among differ-ent populations requires clear definitions of access and methods for measur-ing it. In this study, we evaluate the change in a utility-based park accessibil-ity measure resulting from street conversions in Alameda County, Califor-nia. Our utility-based accessibility measure is constructed from a park activ-ity location choice model we estimate using mobile device data - supplied by StreetLight Data, Inc. - representing trips to parks in that county. The estimated model reveals heterogeneity in inferred affinity for park attributes among different sociodemographic groups. We find, for example, that neigh-borhoods with more lower-income residents and those with more residents of color show a greater preference for park proximity while neighborhoods with higher incomes and those with more white residents show a greater pref-erence for park size and amenities. We then apply this model to examine the accessibility benefits resulting from COVID-19 street conversions to create a set of small park-like open spaces;we find that this policy has improved eq-uity in that marginalized communities including Black, Hispanic, and low-income households receive a disproportionate share of the policy benefits, relative to the population distribution.