Walkable on paper, walkable on the street? Divergence between Walk Score and streetscape quality in Los Angeles
This study of Los Angeles reveals that Walk Score, a widely used accessibility metric, often diverges significantly from actual streetscape quality, systematically underestimating pedestrian disadvantages in lower-income neighborhoods where high paper-based walkability scores mask poor on-the-ground conditions.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
Walking is a fundamental part of city life, yet the tools used to decide where neighborhoods are "walkable" often tell two different stories. For years, urban planners and researchers have relied on a widely used metric called Walk Score to judge how easy it is to get around on foot. This number, which ranges from zero to one hundred, is calculated based on how close a home is to everyday places like grocery stores, schools, and parks, as well as how connected the local street network is. It is a measure of destination accessibility, essentially asking, "How far do I have to travel to get what I need?" However, this calculation happens largely on a map, looking at distances and road layouts from above. It does not account for what a person actually sees and feels while walking down the street. A neighborhood might score highly because it is dense and full of shops, but the sidewalks could be broken, the air might be thick with exhaust, or the streets could be dominated by fast-moving cars. This gap between the map-based score and the ground-level reality is the central question of a new study focused on Los Angeles.
Researchers Jisoo Sim and Narae Lee set out to investigate this disconnect by comparing the official Walk Score with the actual visual quality of the streets in Los Angeles. They analyzed data from 2,232 distinct neighborhoods, known as census block groups, across the city. To get a true picture of the street level, they did not just look at maps; they used thousands of images taken from Google Street View. Using advanced computer vision technology, they trained software to "see" these images just as a human would. The software identified specific elements like sidewalks, trees, buildings, and cars, calculating how much of the view each element occupied. It also assessed more subjective qualities, such as whether a street looked well-maintained, interesting, or inviting versus car-dominated. By combining these visual details into a single measure of street quality, the team could directly compare the "paper" score of walkability with the "street" reality.
The results revealed a surprising and systematic divergence. The study found that a high Walk Score does not guarantee a pleasant walking environment. In fact, the two measures often moved in opposite directions. When the researchers looked at the correlation between the Walk Score and their new street-quality index, they found a negative relationship: generally, the higher the Walk Score, the lower the quality of the streetscape. More than 60 percent of the neighborhoods they studied fell into a discordant category. This means that in the majority of these areas, the map said the neighborhood was walkable, but the street view told a different story, or vice versa.
The study identified two specific types of neighborhoods where this mismatch was most pronounced. One group, which the researchers called "walkable on paper only," consisted of areas with high Walk Scores but poor street conditions. These neighborhoods were characterized by a lack of sidewalks, fewer trees, and a visual dominance of cars. Crucially, these areas were also home to lower-income residents. The average median household income in these "paper-only" neighborhoods was $57,700. In contrast, the neighborhoods that were "walkable on the street only"—meaning they had low Walk Scores but high-quality, pleasant streets—were significantly wealthier, with an average median household income of $124,500. These wealthier areas often had better sidewalks, more greenery, and fewer cars in the view, even though they were less dense and further from daily destinations.
The researchers used statistical models to ensure these patterns were not random. They accounted for factors like population density, land use, and the number of bus stops, and the relationship between income and the type of walkability mismatch remained strong. The data suggests that the neighborhoods with the most urgent need for street improvements—those that are dense and accessible but visually unpleasant and car-heavy—are precisely the ones where lower-income families live. Conversely, wealthier neighborhoods often enjoy high-quality streetscapes even if they are less accessible by walking standards. The study explicitly notes that this does not mean low-income areas lack access to destinations; rather, it highlights that they often bear the burden of poor street conditions despite having good access.
This work challenges the idea that a single number can fully capture the experience of walking in a city. The Walk Score remains a useful tool for understanding how close people are to where they need to go, but it fails to capture the safety, comfort, and beauty of the journey itself. The study suggests that relying solely on accessibility metrics can hide the disadvantages faced by pedestrians in lower-income communities. By looking at both the map and the street, planners and policymakers can identify exactly where improvements are needed. For instance, a neighborhood might not need more shops or shorter blocks, but it might desperately need better sidewalks, more trees, and traffic calming measures to make walking a safe and enjoyable experience. The findings serve as a reminder that true walkability requires attention to both the destination and the path taken to get there.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.