1 Introduction

Parks and other open spaces generate immense value for the public who are able to access them. The City Parks Alliance (2019) categorizes the observed benefits of urban parks as encouraging active lifestyles (Bancroft et al. 2015), contributing to local economies, aiding in stormwater management and flood mitigation, improving local air quality, increasing community engagement (Madzia et al. 2018), and enhancing public equity.

For many, the value of public parks and open public spaces increased during the widespread lockdowns enacted in 2020 to slow the transmission of COVID-19. With other entertainment venues shuttered and people otherwise confined to their homes, periodic use of public space provided an opportunity for physical and emotional relief unavailable in other forms. Paired with this increased demand for public open space — and the with the epidemiological requirement to leave sufficient space between other users — was the related collapse in demand for vehicular travel. As a result, cities around the world began closing select streets to automobile travel, thereby opening them as pedestrian plazas, open streets, or slow streets (Glaser and Krizek 2021; Schlossberg et al. 2021; T. S. Combs and Pardo 2021). The effective result of this policy was to create a number of “parks” in urban areas that may have had poor access previously. Understanding the equitable distribution of these benefits is an important land use policy issue. The potential for non-emergency temporary or permanent street conversions also brings up interesting problems for land use and transportation policy; indeed, the possibility for transportation infrastructure to become a socially beneficial land use that goes beyond serving mobility needs is a tantalizing proposition.

Unfortunately, quantifying the benefits derived from access to parks in general is a complicated problem. Many previous attempts at quantifying access in terms of isochronal distances or open space concentration have resulted in a frustrating lack of clarity on the relationship between measured access and measures of physical and emotional health (Bancroft et al. 2015). Central to this confusion is the fact that people do not always use the nearest park, especially if it does not have qualities that they find attractive. A better methodology would be to evaluate the park activity location choices of people in a metropolitan area to identify which features of parks — distance, amenities, size, etc. — are valued and which are less valued. The resulting activity location choice model would enable the evaluation of utility benefits via the choice model logsum (Handy and Niemeier 1997; Jong et al. 2007).

In this study, we seek to evaluate the socio-spatial distribution of benefits received by residents of Alameda County, California resulting from the temporary conversion of streets to public open spaces during the spring and summer of 2020. We estimate a park activity location choice model using location-based services (LBS) data obtained through StreetLight Data, Inc., a commercial data aggregator. The resulting model illuminates the degree to which simulated individuals living in U.S. Census block groups of varying sociodemographic characteristics value the walking distance between their residence and parks, the size of the parks, and the amenities of parks including sport fields, playgrounds, and walking trails in Alameda county. We then apply this model to examine the inferred monetary benefit resulting from the street conversion policy, and its distribution among different sociodemographic groups.

The paper proceeds in the following manner: A discussion of prior attempts to evaluate park accessibility and preferences is given directly. A Methodology section presents our data gathering and cleaning efforts, the econometric framework for the location choice model, and the approach taken to apply the models to analyze open streets projects. A Results section presents the estimated choice model coefficients alongside a discussion of their implications, including the implied benefits resulting from the street conversions in Alameda County. After presenting limitations and associated avenues for future research, a final Conclusions section outlines the contributions of this study for recreational trip modeling and location choice modeling more generally.