Developing a Supermarket Need Index

Abstract

The New York City Department of City Planning with assistance from the New York City Department of Health and Mental Hygiene developed a supermarket need index to determine the areas in the city with the highest levels of diet-related diseases and largest populations with limited opportunities to purchase fresh foods. The index was created using Geographic Information Systems to measure the need for supermarkets based on high population density, low access to a car at the household level, low household incomes, high rates of diabetes, high rates of obesity, low consumption of fresh fruits and vegetables, low share of fresh food retail, and capacity for new stores. The resulting index identified areas of acute need for additional full-line grocery stores, encompassing portions of the city where approximately three million New Yorkers reside.

Appendix: Data Sources and Tools Used

Data were collected from several sources, including the NYC Department of City Planning, the US Census, The NYC Department of Health and Mental Hygiene, the Bureau of Labor Statistics, and the NY State Department of Agriculture and Markets. Using ArcGIS 9.2 with Spatial Analyst, this information was mapped to reveal diet-related data and land-use conditions across New York City. The boundaries of variables were non-coterminous in some instances, but generally captured similar demographics.

Data and shapefiles came from city and state agencies:

• Supermarkets database – New York State Department of Agriculture and Markets

• Obesity and diabetes – NYC Department of Health and Mental Hygiene

• Consumption of fresh fruits and vegetables – NYC Department of Health and Mental Hygiene

• Share of fresh food retailers to all food retailers – Bureau of Labor Statistics

• Population density – US Census 2000

• Car ownership – US Census 2000

• CDBG eligibility – NYC Department of City Planning

• Geographic shapefiles – NYC Department of City Planning

ESRI ArcGIS 9.2 was used to create the index; generally, the ArcView license with extensions contained the tools needed for this process. An ArcInfo license was required for the Thiessen Polygon analysis.

The following ArcToolbox tools were used:

For specific variable calculations:

• Getis Ord analysis: Geoprocessing tool reference > Spatial Statistics toolbox > Mapping Clusters toolset > Tools – Getis Ord: Use to identify population density and car ownership hot spots, to isolate areas of truly high density and high/low car ownership and not just capture, for example, a single dense building among a more suburban landscape

• Thiessen Polygon creation: Geoprocessing tool reference > Analysis toolbox > Proximity toolset > Tools – Thiessen Polygons: Used to develop approximate trade areas for existing supermarkets in our database – City Planning wanted to capture the areas closest to each existing store and estimate the population living closer to that store than to any other store, in order to determine how well existing stores are able to serve their neighborhood populations.

• Feature to Point conversion: Geoprocessing tool reference > Data Management toolbox > Features toolset > Tools – Feature to point: Used to convert block group polygons (and the population data associated with these polygons) into points, allowing City Planning to capture the residential population within each supermarket trade area.

For general analysis:

• Conversion to Raster

• Spatial Analysis Raster Reclassification

• Raster Calculator

Obesity, diabetes and fruit vegetable consumption variables from Epiquery: NYC Interactive Health Data System – Community Health Survey 2004. Body Mass Index (BMI) is calculated based on respondents’ self-reported weight and height. A BMI between 25.0 and 29.9 is classified as overweight and a BMI of 30 or greater is classified as obese. Diabetes: The exact survey question was: Have you ever been told by a doctor that you have diabetes? Fruit/vegetable consumption: The exact survey question was: How many total servings of fruit and/or vegetables did you eat yesterday? Age-adjusted estimates.