Findings

We looked at 79 BART and Caltrain stations across the Bay Area and measured how many essential amenities (grocery stores, parks, clinics, pharmacies, and childcare) are within walking distance of each one. Here is what we found.

What the data shows

The charts below summarize the core-peripheral gap from three angles: overall amenity counts, which stations have the most urgent unmet need, and how the gap breaks down by amenity type. For the full statistical tests behind these patterns, see Methods & Results.

What this shows. Figure 1 puts both groups side by side, and the gap is clear. Some core stations have 50+ amenities nearby. Others, like Coliseum and Bay Fair, have 3 and 5.

Figure 1. Distribution of total walkable amenities (½-mile radius) by station type. Each point is one station. Data: OpenStreetMap amenity extracts, FY2025 BART and Caltrain ridership, n=79 stations.

What this shows. Figure 2 ranks stations by how much their car-free population exceeds what the neighborhood can offer on foot. Coliseum, Bay Fair, and MacArthur are all high-ridership core stations by the numbers, but they look peripheral when you count what's around them.

Figure 2. Top 15 stations ranked by unmet need index. Blue = core, orange = peripheral. Index = % zero-vehicle households × (1 − normalized amenity count), min-max normalized across all 79 stations. Data: ACS 2024 5-year estimates, OpenStreetMap.

What this shows. Figure 3 breaks the gap down by category, and the clinic numbers are the ones that matter most for daily life. At peripheral stations, the average rider is within walking distance of less than one clinic. At core stations, that number is 2.5. Pharmacies follow the same pattern.

Note on category selection. The chart below shows the five amenity categories with the largest and most consistent gaps after FDR correction. Convenience stores showed the largest raw effect (Glass Delta = 2.47) but were excluded here because convenience store counts are heavily skewed by a few downtown SF stations, making the mean misleading as a summary. The full 10-category breakdown is in Methods & Results.

Figure 3. Mean amenity count within ½ mile by category and station type. Hospitals, convenience stores, doctors, and kindergartens are excluded from this summary chart because their permutation test results were not significant after FDR correction. Data: OpenStreetMap, n=40 core, n=39 peripheral.

What this shows. Figure 4 maps out the relationship between car-free households and amenity access directly — having more car-free households nearby tends to mean more amenities, but that breaks down for several East Oakland stations. Coliseum sits at 29% car-free households with only 3 amenities nearby. The trendline goes up; those stations go sideways.

Figure 4. % zero-vehicle households vs. total walkable amenities. Gray line = OLS trend (Spearman rho = +0.476, p < 0.001). Labeled stations are in the top decile of unmet need. Data: ACS 2024, OpenStreetMap.

Unmet need across the network

The map below shows unmet need index at each station, colored from light to dark. Darker = more urgent. Stations with a black outline are in the top quartile. These are the places where the combination of low amenity access and high car-free household rates is most severe. Use this alongside the Station Access Explorer map to get a sense of where the network is failing its most transit-dependent riders geographically.

Three-way equity view

The figure below plots each station across three dimensions at once: car-free household rate, non-white resident share, and total amenities. Color shows unmet need index. Stations in the dark red corner (high on both demographic axes, low on amenities) are where the equity gap is most acute. Rotate the chart to explore the relationship from different angles.