Food Desert Transit Access and Persistence

ISEF Category: Behavioral and Social Sciences

Subcategory: Sociology and Anthropology  ·  Difficulty: Advanced  ·  Setup: Home Setup  ·  Time: Full Year

The Hook

A new bus line does not always mean better grocery access. Some neighborhoods still sit far from healthy food, even after transit changes. You can measure that gap with maps, census data, and graph-based distance scores. This project asks whether transit actually moved the needle, or only made the map look better.

What Is It?

Food deserts are places where reaching a grocery store takes too much time, money, or effort. In this project, you track whether those places stayed stuck from 2010 to 2024, even when new transit lines opened.

Think of a city like a web of dots and lines. The dots are neighborhoods, stops, and stores, and the lines are roads and transit routes. A graph-based accessibility metric turns that web into a number, so you can compare how hard it is to reach food by transit instead of just drawing circles on a map. GTFS feeds are the schedule files that tell you when buses and trains actually run. ACS, the American Community Survey, gives you neighborhood traits like income and car ownership that may help explain who gained access and who did not.

Why This Is a Good Topic

This is a strong science fair topic because you can test a real public policy question with public data. You can compare neighborhoods, transit changes, and food access with clear metrics. You will learn GIS, graph analysis, data cleaning, and how to argue from evidence instead of opinion.

Research Questions

• How does adding a new transit line change effective food-access distance for nearby neighborhoods?
• What is the effect of transit service frequency on food-access distance after a route opens?
• Does food access improve more in low-income neighborhoods than in higher-income neighborhoods after transit changes?
• To what extent do neighborhoods with lower car ownership see larger access gains from new transit lines?
• Which graph-based accessibility metric best predicts whether a neighborhood gains food access after transit expansion?
• How does distance to the nearest full-service grocery store change when you measure network travel instead of straight-line distance?

Basic Materials

• A laptop or desktop computer with internet access.
• Spreadsheet software such as Google Sheets or Excel.
• QGIS for map cleaning and layer joins.
• USDA Food Access Research Atlas data.
• American Community Survey five-year tables from the U.S. Census Bureau.
• GTFS feeds from local transit agencies or Transitland.
• U.S. Census tract boundary files.

Advanced Materials

• Python with pandas, GeoPandas, NetworkX, and OSMnx.
• QGIS or ArcGIS Pro for spatial joins and map checks.
• Historical GTFS archives from transit agencies or MobilityData.
• U.S. Census TIGER/Line shapefiles for roads and tracts.
• USDA Food Access Research Atlas extracts for multiple years.
• ACS five-year tract tables and comparison variables.
• A spatial database such as PostGIS.

Software & Tools

• QGIS: Cleans spatial layers, checks boundaries, and maps access patterns.
• Python: Joins census, transit, and food access data, then runs the metrics.
• GeoPandas: Handles tract shapefiles and spatial joins.
• NetworkX: Builds graph routes and calculates accessibility scores.
• Jupyter Notebook: Keeps code, notes, and figures in one place.

Experiment Steps

1. Define the neighborhood unit you will compare, such as census tracts or block groups.
2. Choose the food-access measure you will treat as your main outcome, then state how you will calculate it.
3. Build a before-and-after transit comparison that separates neighborhoods with new service from similar neighborhoods without it.
4. Plan controls for car ownership, income, density, and grocery store count so transit effects do not get blurred.
5. Pick one statistical test or model that matches your data structure, then decide how you will report effect size and uncertainty.

Common Pitfalls

• Using straight-line distance instead of network travel, which hides barriers like rivers, highways, and disconnected streets.
• Mixing transit feeds from different dates, which makes a route look new or removed when the schedule just changed.
• Comparing census tracts with block groups without re-scaling the data, which breaks the access metric.
• Treating every new transit line as a true access gain, even when service runs too rarely to help grocery trips.
• Ignoring grocery store openings and closings, which can make a transit change look more powerful than it was.

What Makes This Competitive

A stronger version of this project does more than map hotspots. It compares matched neighborhoods, uses a real before-and-after design, and tests whether transit changes beat a no-change control group. You can push it further by adding uncertainty estimates, sensitivity checks for different access metrics, and a city-to-city comparison. That turns a simple map into a claim about how transit changes food access.

Project Variations

• Focus on one metro area and compare bus-only upgrades with rail extensions.
• Swap grocery stores for farmers markets or healthy corner stores to test whether the access pattern changes.
• Compare straight-line distance, drive-time, and transit-based accessibility to see which measure best matches neighborhood food access.

Learn More

• USDA Food Access Research Atlas: Find tract-level food access maps and data on the USDA Economic Research Service site.
• American Community Survey: Get neighborhood income, car ownership, and household data on the U.S. Census Bureau site.
• Census TIGER/Line Shapefiles: Download tract and boundary files from the U.S. Census Bureau for map building.
• MobilityData GTFS Reference: Read the GTFS feed format and transit calendar rules on the MobilityData site.
• QGIS Documentation: Learn how to join layers, clip maps, and make clean accessibility figures on the QGIS site.
• NetworkX Documentation: Look up shortest-path and centrality tools for graph-based accessibility on the NetworkX site.