Measuring inequalities in urban systems: An approach for evaluating the distribution of amenities and burdens

T.M.Logan, M.J.Anderson, T.G.Williams, L.Conrow

A public repository used in the creation of the recent publication, Measuring inequalities in urban systems: An approach for evaluating the distribution of amenities and burdens

Please cite as: Logan, T., Anderson, M., Williams, T., and Conrow, L. “Measuring Inequality in the built environment: an approach for evaluating the distribution of amenities and burdens”. Computers, Environment and Urban Systems.

Overview

In this publication, we analysed the inequality of network-based accessibility to supermarkets in 10 major US cities. We did this by utilising open-source data for supermarket locations and census data at a census block level. Technically, we leveraged our previous work on OSRM Network Query to calculate the nearest network distance and Inequalipy: Measuring inequality among distributions to evaluate the inequality among the distance distribution. This repository is used as a demonstration of just one example where inequalipy can be used within the urban environment. Other examples include distributions of both "goods" (where a high value is typically favourable, e.g. income) and "bads" (where a low value is typically favourable, e.g. exposure to a natural hazard).

Utilising Query Access OSRM

Using OSRM and Python we calculated the nearest distance from residents homes to their nearest supermarket. This blog post describes this process.

Utilising Inequalipy

As part of this publication, we developed a Python and R package to quickly and efficiently calculate inequality among distributions. This package provides functions for the following:

• Kolm-Pollak Equally-Distributed Equivalent (EDE) and Index
• Atkinson EDE and Index
• Gini Index