Geography
Manhattan only (county boundary polyfill -> H3 R8)
← Back to use cases
Use case
311 Health & Sanitation Intelligence (Manhattan Demo) - Powered by Etherdata's Canonical US Census at H3 Resolution 8
Etherdata.ai publishes canonical, analysis-ready spatial datasets so teams can move from interesting maps to defensible decisions. This page demonstrates a simple but transformative principle: operational signals become decision-grade only when they are anchored to canonical denominators.
We use NYC 311 service requests as the operational signal and Etherdata's canonical census layer as the backbone. The result is a monitoring view that answers a question counts alone cannot: where are health and sanitation issues truly over-indexing relative to the number of people exposed?
Geography
Manhattan, New York (H3 Resolution 8)
Operational signal
NYC 311 health & sanitation service requests
Backbone
Etherdata canonical census denominators
Focus
Complaint rate, operational friction, risk index
Why the census table creates the value
A raw 311 request count is not comparable across neighborhoods. It is a mixture of: (1) exposure (how many people live there), (2) urban form (housing density, housing stock), and (3) reporting dynamics (language access, time, trust, civic habits).
Etherdata's canonical census table turns 311 from a volume map into a comparable measurement layer by providing:
- Denominators (population, households, housing units) to convert counts into stable rates
- Equity context (e.g., rent burden) to interpret who is likely most constrained
- Join reliability via deterministic H3 indexing (no hand-built spatial joins, no drifting boundaries)
This is the core value proposition: the census layer is not another dataset. It is the scaffold that makes every other dataset more interpretable, more comparable, and more actionable.
Scope and structure
Unit of analysis
H3 resolution 8 hexagons (one row per cell)
Signal
NYC 311 service requests (health & sanitation taxonomy defined explicitly)
Backbone
Etherdata canonical census (denominators + context)
Charts
The charts below are placeholders for the current Looker Studio report outputs (4 charts). They are intentionally specified with exact field bindings so they can be embedded cleanly on the public site.
Health & Sanitation Complaint Rate
This is the primary monitoring map: health requests per 1,000 population. It is the most defensible hotspot view because it removes the dominant confounder (population exposure). In the report output, a small number of cells reach the high end of the scale (near ~561 per 1k), standing apart from the broader baseline across Manhattan.
Decision use
Nominate hotspot clusters for investigation and response planning.
Why census is essential
The denominator is what makes high meaningful and comparable.
Health & Sanitation Risk Index
The risk index is where the census layer becomes explicitly transformative. It blends three standardized components into a transparent triage score: complaint rate intensity, rent burden (vulnerability), and average time-to-close (operational friction). In the report output, the index spans roughly -2.57 to 0.64, producing a sharper prioritization map than rates alone.
Decision use
Allocate limited remediation and inspection capacity to the highest-priority cells.
Why census is essential
Vulnerability context makes prioritization more equitable and defensible.
Complaint Volume vs Population
This chart demonstrates the trap of raw counts. As population increases, complaint volume increases - predictably. The purpose is not discovery; it is explainability: it shows stakeholders why most complaints is not the same as most at-risk. Outliers above the general relationship indicate cells generating more complaints than expected for their exposure.
Decision use
Defend rate-based prioritization over count-based prioritization.
Why census is essential
Population is the exposure baseline that makes the scatter interpretable.
Equity Lens: Rent Burden vs Complaint Rate
This chart makes the equity argument legible. It asks: do elevated health and sanitation complaint rates concentrate where households are more financially constrained? In the output, most cells cluster in a rent burden band (~20-35 percent) with modest complaint rates, while a small set of outliers reach very high rates (approaching the top of the axis near ~600). These outliers are the look closer cells.
Decision use
Defend rate-based prioritization over count-based prioritization.
Why census is essential
Population is the exposure baseline that makes the scatter interpretable.
Conclusion
This Manhattan demo is intentionally simple - and that is the point. The lift is not making a map. The lift is establishing a canonical backbone that makes operational data comparable, explainable, and fair.
With Etherdata's canonical census table, NYC 311 becomes a reusable health and sanitation intelligence layer: it supports hotspot detection, operational triage, and equity-aware prioritization in a single framework.