Every indicator, source, crosswalk, and reconciliation used to produce the Neighborhood Explorer.
Every source contributing to the Neighborhood Explorer. Green = fresh, amber = stale beyond its window, red = degraded or failed, gray = not yet loaded.
| Source | Agency | Dataset | Cadence | Stale window | Last refreshed | Rows | Status |
|---|---|---|---|---|---|---|---|
2020 CDTA Boundaries dcp_cdta_boundaries | DCP | xn3r-zk6y |
| 365 d |
| 400 d |
| Jul 17, 2026 |
| 71 |
| fresh |
2020 NTA Boundaries dcp_nta_boundaries | DCP | 9nt8-h7nd | 365 d | 400 d | Jul 17, 2026 | 262 | fresh |
2020 Tract Equivalency dcp_tract_equivalency | DCP | hm78-6dwm | 1095 d | 1200 d | Jul 17, 2026 | 2,327 | fresh |
ACS 5-year (Census API) census_acs5 | US Census Bureau | acs5 | 365 d | 400 d | Jul 17, 2026 | 41,886 | fresh |
CDC PLACES (NYC counties) cdc_places_nyc | CDC | cwsq-ngmh | 365 d | 400 d | Jul 17, 2026 | 89,238 | fresh |
Community Districts dcp_community_districts | DCP | 5crt-au7u | 365 d | 400 d | Jul 17, 2026 | 71 | fresh |
DOHMH Community Health Profiles dohmh_chp | DOHMH | nyc_gov_landing | 365 d | 425 d | Jul 17, 2026 | 4,538 | fresh |
DYCD Contracts dycd_contracts | DYCD | graj-69em | 30 d | 45 d | Jul 17, 2026 | 12,208 | fresh |
DYCD NDA Boundaries dycd_nda_boundaries | DYCD | vd7c-qjsx | 365 d | 400 d | Jul 17, 2026 | 41 | fresh |
DYCD Program Sites dycd_program_sites | DYCD | ebkm-iyma | 30 d | 45 d | Jul 17, 2026 | 28,830 | fresh |
DYCD site-contract join rollups derive_site_contract_rollups | internal | derived | 30 d | 45 d | Jul 17, 2026 | 28,830 | fresh |
Long-format indicator values derive_indicators | internal | derived | 30 d | 45 d | Jul 17, 2026 | 5,445 | fresh |
NDA code reconstruction derive_nda_code_map | internal | derived | 365 d | 400 d | Jul 17, 2026 | 41 | fresh |
NYCgov Poverty Measure nycgov_poverty | NYC Mayor's Office of Opportunity | cts7-vksw | 365 d | 425 d | Jul 17, 2026 | 12 | fresh |
NYPD Complaints Historic nypd_historic | NYPD | qgea-i56i | 30 d | 45 d | Jul 17, 2026 | 20,868 | fresh |
NYPD Complaints YTD nypd_ytd | NYPD | 5uac-w243 | 30 d | 45 d | Jul 17, 2026 | 3,979 | fresh |
Police Precinct Boundaries nypd_precinct_boundaries | DCP | y76i-bdw7 | 365 d | 400 d | Jul 17, 2026 | 78 | fresh |
Precinct-NDA overlap (>= 5%) derive_precinct_nda_overlap | internal | derived | 365 d | 400 d | Jul 17, 2026 | 153 | degraded |
Tract/CD -> NDA/Precinct crosswalks derive_crosswalks | internal | derived | 30 d | 60 d | Jul 17, 2026 | 4,167 | fresh |
Every active indicator, its native geography, aggregation method, and provenance.
Indicator geographies (tract, NTA, CD, precinct) do not align perfectly with NDA borders. The Neighborhood Explorer uses area-weighted intersection crosswalks; residuals are reported in the reconciliation section.
Crime indicators are precinct-native. The Neighborhood Explorer aggregates complaints from NYPD YTD (5uac-w243) and NYPD Historic (qgea-i56i) into monthly counts by precinct × law category. Complaint records themselves are not stored; only the aggregated counts.
DYCD publishes two related datasets: the program-site directory (site ID as primary key) and the contracts ledger (contract number as primary key). To attribute contract dollars to an NDA, the pipeline joins on the contract-number suffix stamped on each site's program record. A site is matched when the suffix resolves to at least one contract row and unmatched when it does not.
Unmatched sites are still counted in dycd_site_count; their funding is excluded from dycd_contract_dollars_totaland shown separately in the KPI ribbon as "X of Y sites matched". Unmatched rates greater than 10% are surfaced with an inline caveat on the NDA profile.
Program sites are bucketed into three cohorts based on the raw NDA field on the source record:
| Check | Value | Notes |
|---|---|---|
| NDA code count | 41 | Expect 41. |
| Out-of-NDA program sites | 9,211 | Sites geolocated in NYC but not inside any of the 41 NDA polygons. |
| NDA missing on source | 52 | Source row lacked a valid NDA value. |
| Contract join quality | 25,416 matched / 3,414 unmatched | 11.8% of program sites are unlinked to a contract number. |
| ACS citywide sum · 2020-2024 | 8,483,844 | Sum of B01003_001E across every NYC tract, from the ACS 5-year API. |
| NDA-covered population (area-weighted) | 4,901,867 | Sum of tract population × NDA area-weight across all 41 NDAs. NDA coverage 57.8% of NYC. |
| Reconciliation delta | -42.22% | Expected: NDA-covered ≤ citywide (NDAs cover roughly 55–60% of NYC). Validated Jul 17, 2026. Tolerance ±0.5% applies to the ACS ingest sum, not to the NDA subset. |
| ACS vintage | 2020-2024 | Cached in neighborhoods_meta.current_acs_vintage. |
Indicator geographies (tract, NTA, CD, precinct) do not align perfectly with NDA borders. Tract-weighted NDA aggregation is used for ACS and CDC PLACES; area-weighted CD-to-NDA mapping is used for DOHMH profiles. The methodology page documents the exact weights and a citywide reconciliation delta from DATA_REPORT.md.
Affected indicator families: Demographics, Poverty, Health
Workaround: Show reconciliation delta and per-tract residual on the methodology page.
DetailsDYCD internal community-needs-assessment data and participant-level program data are not public. Profiles use public indicators until DYCD data connects inside the firewall.
Affected indicator families: DYCD Footprint, Contract Funding
Workaround: Use DYCD Program Sites and DYCD Contracts as the public footprint until DYCD internal data connects.
DetailsFaith institutions inventory, broader CBO inventory, and other non-DYCD community assets are roadmap. v1's asset layer is DYCD's own program footprint from ebkm-iyma.
Affected indicator families: DYCD Footprint
Workaround: Phase 2 asset table on the methodology page documents the roadmap.
DetailsCandidate datasets that would broaden the community-asset layer beyond DYCD sites.
| Candidate | Owner agency | Public status | Dataset ID | Feasibility | Deferral rationale |
|---|---|---|---|---|---|
| Faith institutions inventory | Various | Partial (some via ACS, some non-public) | - | Requires a curated NYC-wide inventory beyond the ACS Religious Congregations and Membership Study coverage; DYCD does not publish one. | v1 anchors on DYCD's own program footprint; broader community-asset framing is a roadmap layer per DOSSIER. |
| Broader CBO / nonprofit inventory | Various (NY AG, IRS, NYC MOCS) | Partial | IRS 990 filings; NYC MOCS awarded contracts | Requires cross-agency reconciliation of nonprofit registration, address, and program category. | Would be its own product; keeping neighborhoods focused on DYCD-anchored view. |
| Public library branches | NYPL, BPL, QPL | Yes | Multiple, per system | Three separate library systems publish separately; a unified branch dataset would be a valuable community-asset layer. | v1 non-goal; adds value in Phase 2 as a general community-asset layer. |
| NYC parks and public open space | NYC Parks | Yes | Socrata (multiple datasets) | Well-published; a natural next community-asset layer. | v1 non-goal; keeps scope tight to DYCD footprint. |