Projects

Quality of Life of NYC Children

Pythonscikit-learnGeoPandasGISMapping
Fall 2021
Intro to Urban Data Informatics
Columbia GSAPP
Prof. Boyeong Hong

Team:
Kirthi Balakrishnan,
Shreya Arora,
Lizzie Lee,
Christian Budow

A composite Quality of Life Index (0–10) for children across NYC's 55 PUMAs, integrating infrastructure access, economic conditions, environmental quality, and social demographics into a single comparable framework.

Scroll
01

The Question

Method
Composite index
4 domains, 0–10 scale
55 NYC PUMAs

NYC Open Data
2019 ACS 5-Year

Existing quality-of-life indices treat children as a side concern. Yet how kids interact with the built environment — whether they can walk to a library, breathe clean air at school, or live in a household above the rent-burden line — shapes outcomes that ripple across decades.

This project builds a composite Quality of Life Index for children across all 55 of NYC's Public Use Microdata Areas, drawing on the UNICEF Child Friendly Cities Initiative, Arup's Cities Alive: Designing for Urban Childhoods report, and ONE NYC 2050's well-being indicators — and adapting them to the lived geography of New York's five boroughs.

02

Approach

Four Domains
Domains
Access
Economic
Environmental
Social

Each domain rolls multiple indicators into a normalized sub-score, then the four are summed and re-scaled to produce a single 0–10 composite. Indicators where lower values are better (PM2.5, rent burden) are inverse-scored so the index reads consistently.

  • Access. Density of bus stops, subway stations, healthcare facilities, libraries, and parks per PUMA — NYC Open Data shapefiles.
  • Economic. Median household income and rent burden (gross rent as % of income) — 2019 ACS 5-Year Estimates. Rent burden ≥ 35% scores 1.
  • Environmental. Fine particulate matter (PM2.5) and water quality (chlorine, turbidity, fluoride) — NYC Open Data Portal APIs.
  • Social. Population under 18, age-group breakdowns, and race/ethnicity composition — 2019 ACS 5-Year Estimates.
Tools & Libraries

Python, GeoPandas, scikit-learn (MinMaxScaler), pandas, NumPy, Plotly, Folium, spatial joins on PUMA boundaries.

03

Outputs

Composite index and component maps
Composite Quality of Life Index choropleth across NYC PUMAs
COMPOSITE INDEX
Final 0–10 QoL score by PUMA — Viridis scale, summed across four domains.
Component maps and demographic breakdowns
COMPONENTS
Median income, rent burden, and infrastructure density heatmaps that feed the composite.
04

Findings

  • Quality of life for children is not legible from any single metric. The PUMAs that score highest on infrastructure access do not consistently score highest on environmental or economic dimensions — a single composite is necessary to capture the full picture.
  • Economic and environmental burdens cluster. Rent-burdened PUMAs and high-PM2.5 PUMAs overlap in ways that compound the disadvantage for children growing up in those neighborhoods.
  • Existing global frameworks need NYC-specific tuning. UNICEF and Arup's frameworks are useful starting points, but their indicator weightings don't reflect the realities of dense, vertical, transit-rich cities like New York. Tuning the composite to NYC's geography surfaces patterns the global frameworks obscure.