2022 to 2025, with January to June 2026 held out · 15,610,022 calls from the public, 2022 to June 2026 · Martin Ngoh
Urgent 911 calls in New York wait about the same across neighborhoods. Less urgent calls don't.
Critical calls: an adjusted median of 5.0 minutes in the lowest-income fifth of neighborhoods and 4.1 in the highest; 4.9 in majority-Black, 5.0 in majority-Hispanic and 4.2 in majority-White neighborhoods. Gaps of 0.6 to 0.8 minutes, under the 1-minute bar (intervals reach about 1.1). Serious calls: about 0.9 minutes, under the 20% bar.
Less urgent calls: 42%, 28% and 40% longer for non-critical crimes in progress (lowest-income, majority-Black, majority-Hispanic); 69%, 45% and 79% for calls not about a crime in progress. All above the 20% bar.
Call type, hour, precinct workload, distance to the station house and officers per call barely move the gaps. Inside the same precinct they disappear, which locates them between precincts without explaining them. The majority-Black gaps were smaller in 2026; the others held.
Median minutes from the call being entered to the first officer arriving, calls from the public, 2022 to 2025, with each group's calls reweighted to the citywide mix of call types, hours of the week and precinct busyness. Gaps against highest-income fifth neighborhoods with 95% intervals.
Median minutes from the call being entered to the first officer arriving, calls from the public, 2022 to 2025, with each group's calls reweighted to the citywide mix of call types, hours of the week and precinct busyness. Gaps against majority-White neighborhoods with 95% intervals.
Regression of log minutes with errors clustered by neighborhood: the percent difference in typical minutes after each set of controls, added one at a time. Officers per call uses a 2026 staffing snapshot. "Same precinct only" compares neighborhoods inside the same precinct.
| Neighborhoods | Priority | No controls | Call type | + hour of week | + precinct busyness | + distance to station | + officers per call | Same precinct only |
|---|---|---|---|---|---|---|---|---|
| Lowest-income fifth | Critical | +27% | +28% | +27% | +27% | +28% | +30% | -8% |
| Lowest-income fifth | Serious | +20% | +22% | +23% | +23% | +22% | +28% | -10% |
| Lowest-income fifth | Non-critical | +54% | +48% | +50% | +50% | +50% | +57% | -7% |
| Lowest-income fifth | Not a crime in progress | +66% | +60% | +61% | +61% | +64% | +73% | -11% |
| Majority-Black | Critical | +22% | +21% | +20% | +20% | +18% | +23% | -3% |
| Majority-Black | Serious | +18% | +19% | +19% | +19% | +16% | +26% | -2% |
| Majority-Black | Non-critical | +38% | +32% | +32% | +32% | +29% | +43% | -2% |
| Majority-Black | Not a crime in progress | +44% | +41% | +42% | +42% | +37% | +57% | -5% |
| Majority-Hispanic | Critical | +23% | +25% | +25% | +25% | +25% | +28% | -6% |
| Majority-Hispanic | Serious | +18% | +20% | +21% | +21% | +21% | +27% | -10% |
| Majority-Hispanic | Non-critical | +50% | +45% | +47% | +47% | +47% | +56% | -8% |
| Majority-Hispanic | Not a crime in progress | +65% | +61% | +64% | +64% | +66% | +78% | -7% |
| Majority-Asian | Critical | +9% | +7% | +7% | +7% | +8% | +10% | -10% |
| Majority-Asian | Serious | +11% | +14% | +13% | +13% | +15% | +18% | -13% |
| Majority-Asian | Non-critical | +41% | +36% | +35% | +35% | +38% | +42% | -7% |
| Majority-Asian | Not a crime in progress | +66% | +47% | +47% | +47% | +52% | +57% | -12% |
| No majority | Critical | +6% | +6% | +6% | +6% | +5% | +8% | -5% |
| No majority | Serious | +0% | +1% | +1% | +1% | +1% | +6% | -7% |
| No majority | Non-critical | +8% | +6% | +6% | +6% | +5% | +11% | -6% |
| No majority | Not a crime in progress | +11% | +10% | +10% | +10% | +10% | +17% | -9% |
The 20 comparisons set in the plan before the analysis ran, each against the highest-income fifth or majority-White neighborhoods. Adjusted median minutes; gap with a 95% interval from resampling neighborhoods; p adjusted for all of them (Holm). A gap "matters" if it is clear and at least the smallest gap that matters, set in advance. It is "smaller than the smallest gap that matters" if the whole interval sits inside that bound. * Fewer than 10 neighborhoods: reported, not headlined.
| Priority | Neighborhoods | Median (min) | Gap (min) | Gap | p | Bar | Result |
|---|---|---|---|---|---|---|---|
| Critical | Lowest-income fifth | 5.0 vs 4.1 | +0.8 (+0.5 to +1.1) | +20% | 2.8e-06 | 1 minutes | real but smaller than the smallest gap that matters |
| Serious | Lowest-income fifth | 6.9 vs 5.9 | +0.9 (+0.5 to +1.5) | +16% | 0.0037 | 20% | real but smaller than the smallest gap that matters |
| Non-critical | Lowest-income fifth | 13.5 vs 9.5 | +4.0 (+2.5 to +5.4) | +42% | 1.6e-06 | 20% | gap that matters |
| Not a crime in progress | Lowest-income fifth | 32.5 vs 19.2 | +13.3 (+8.5 to +19.0) | +69% | 7.3e-06 | 20% | gap that matters |
| Critical | Majority-Black | 4.9 vs 4.2 | +0.6 (+0.3 to +1.0) | +15% | 0.006 | 1 minutes | smaller than the smallest gap that matters |
| Critical | Majority-Hispanic | 5.0 vs 4.2 | +0.7 (+0.4 to +1.1) | +17% | 0.00095 | 1 minutes | real but smaller than the smallest gap that matters |
| Critical | Majority-Asian * | 4.5 vs 4.2 | +0.2 (-0.1 to +0.9) | +5% | 0.8 | 1 minutes | smaller than the smallest gap that matters |
| Critical | No majority | 4.4 vs 4.2 | +0.2 (-0.1 to +0.4) | +5% | 0.65 | 1 minutes | smaller than the smallest gap that matters |
| Serious | Majority-Black | 6.8 vs 6.0 | +0.9 (+0.3 to +1.5) | +14% | 0.041 | 20% | real but smaller than the smallest gap that matters |
| Serious | Majority-Hispanic | 6.8 vs 6.0 | +0.9 (+0.4 to +1.4) | +14% | 0.0092 | 20% | real but smaller than the smallest gap that matters |
| Serious | Majority-Asian * | 6.9 vs 6.0 | +1.0 (+0.2 to +1.8) | +16% | 0.07 | 20% | inconclusive |
| Serious | No majority | 6.0 vs 6.0 | +0.1 (-0.3 to +0.4) | +1% | 0.8 | 20% | smaller than the smallest gap that matters |
| Non-critical | Majority-Black | 12.3 vs 9.6 | +2.7 (+1.4 to +4.2) | +28% | 0.003 | 20% | gap that matters |
| Non-critical | Majority-Hispanic | 13.4 vs 9.6 | +3.9 (+2.3 to +5.5) | +40% | 8.4e-05 | 20% | gap that matters |
| Non-critical | Majority-Asian * | 12.7 vs 9.6 | +3.1 (+1.6 to +6.0) | +33% | 0.023 | 20% | gap that matters |
| Non-critical | No majority | 10.0 vs 9.6 | +0.4 (-0.2 to +1.1) | +4% | 0.8 | 20% | smaller than the smallest gap that matters |
| Not a crime in progress | Majority-Black | 30.4 vs 21.0 | +9.4 (+5.2 to +14.3) | +45% | 0.00075 | 20% | gap that matters |
| Not a crime in progress | Majority-Hispanic | 37.6 vs 21.0 | +16.6 (+12.0 to +21.9) | +79% | 5.9e-10 | 20% | gap that matters |
| Not a crime in progress | Majority-Asian * | 30.6 vs 21.0 | +9.6 (+0.0 to +23.6) | +46% | 0.65 | 20% | inconclusive |
| Not a crime in progress | No majority | 22.9 vs 21.0 | +1.9 (-1.0 to +4.8) | +9% | 0.8 | 20% | inconclusive |
Gap in adjusted median minutes for each part of the wait: from entry to a unit being dispatched, and from dispatch to arrival.
| Neighborhoods | Priority | Entry to dispatch | Dispatch to arrival |
|---|---|---|---|
| Lowest-income fifth | Critical | +0.3 (+0.2 to +0.3) | +0.4 (+0.1 to +0.7) |
| Lowest-income fifth | Serious | +0.4 (+0.3 to +0.5) | +0.2 (-0.2 to +0.6) |
| Lowest-income fifth | Non-critical | +1.0 (+0.7 to +1.4) | +1.3 (+0.6 to +2.1) |
| Lowest-income fifth | Not a crime in progress | +1.4 (+0.9 to +2.1) | +7.9 (+4.9 to +11.5) |
| Majority-Black | Critical | +0.1 (+0.1 to +0.2) | +0.5 (+0.2 to +0.8) |
| Majority-Black | Serious | +0.2 (+0.1 to +0.3) | +0.5 (+0.1 to +0.9) |
| Majority-Black | Non-critical | +0.6 (+0.3 to +0.9) | +1.2 (+0.3 to +1.9) |
| Majority-Black | Not a crime in progress | +0.9 (+0.5 to +1.4) | +6.2 (+3.2 to +9.6) |
| Majority-Hispanic | Critical | +0.3 (+0.2 to +0.4) | +0.3 (+0.0 to +0.6) |
| Majority-Hispanic | Serious | +0.4 (+0.2 to +0.5) | +0.2 (-0.2 to +0.5) |
| Majority-Hispanic | Non-critical | +1.1 (+0.7 to +1.5) | +1.2 (+0.4 to +1.9) |
| Majority-Hispanic | Not a crime in progress | +1.6 (+1.1 to +2.3) | +10.4 (+7.6 to +13.4) |
| Majority-Asian | Critical | +0.0 (-0.0 to +0.1) | +0.2 (-0.1 to +0.7) |
| Majority-Asian | Serious | +0.2 (+0.1 to +0.2) | +0.7 (+0.1 to +1.6) |
| Majority-Asian | Non-critical | +0.5 (+0.0 to +0.9) | +1.8 (+1.0 to +3.6) |
| Majority-Asian | Not a crime in progress | +1.0 (+0.1 to +1.9) | +5.4 (-0.5 to +15.8) |
| No majority | Critical | +0.0 (-0.0 to +0.1) | +0.2 (-0.1 to +0.4) |
| No majority | Serious | +0.0 (-0.0 to +0.1) | +0.0 (-0.2 to +0.3) |
| No majority | Non-critical | +0.0 (-0.1 to +0.2) | +0.2 (-0.3 to +0.6) |
| No majority | Not a crime in progress | +0.2 (-0.1 to +0.4) | +1.2 (-1.0 to +3.4) |
Share of calls from the public with no recorded arrival, which the waits above leave out. Counting them as waiting until the event was closed makes the gaps for less urgent calls larger, not smaller.
| Neighborhoods | Critical | Serious | Non-critical | Not a crime in progress |
|---|---|---|---|---|
| Lowest-income fifth | 4.3% | 6.4% | 17.4% | 45.0% |
| Second fifth | 3.9% | 4.8% | 12.3% | 37.7% |
| Middle fifth | 3.0% | 4.8% | 11.2% | 34.6% |
| Fourth fifth | 3.8% | 5.5% | 10.4% | 35.7% |
| Highest-income fifth | 3.1% | 7.5% | 17.3% | 42.4% |
| Majority-Black | 4.4% | 5.4% | 13.1% | 36.8% |
| Majority-Hispanic | 4.3% | 6.0% | 16.4% | 44.8% |
| Majority-Asian | 1.8% | 3.7% | 11.4% | 34.0% |
| No majority | 3.3% | 5.3% | 11.8% | 35.9% |
| Majority-White | 3.0% | 7.4% | 17.1% | 43.2% |
Gap in adjusted median minutes in each half of the window and in January to June 2026, which was set aside as a check.
| Priority | Neighborhoods | 2022-23 | 2024-25 | 2026 holdout |
|---|---|---|---|---|
| Critical | Lowest-income fifth | +0.8 (+0.5 to +1.1) | +0.9 (+0.5 to +1.2) | +1.0 (+0.5 to +1.3) |
| Serious | Lowest-income fifth | +0.9 (+0.4 to +1.5) | +0.9 (+0.5 to +1.5) | +1.1 (+0.5 to +1.6) |
| Non-critical | Lowest-income fifth | +3.2 (+2.0 to +4.6) | +4.8 (+3.0 to +6.7) | +6.3 (+4.0 to +9.1) |
| Not a crime in progress | Lowest-income fifth | +12.0 (+7.5 to +17.3) | +14.7 (+9.7 to +21.2) | +14.3 (+9.2 to +21.1) |
| Critical | Majority-Black | +0.6 (+0.3 to +1.0) | +0.6 (+0.2 to +1.0) | +0.2 (-0.2 to +0.6) |
| Critical | Majority-Hispanic | +0.7 (+0.4 to +1.0) | +0.7 (+0.3 to +1.1) | +0.8 (+0.3 to +1.2) |
| Critical | No majority | +0.2 (-0.1 to +0.4) | +0.2 (-0.1 to +0.4) | -0.0 (-0.4 to +0.3) |
| Serious | Majority-Black | +0.9 (+0.3 to +1.6) | +0.8 (+0.3 to +1.5) | +0.4 (-0.2 to +1.0) |
| Serious | Majority-Hispanic | +0.9 (+0.3 to +1.4) | +0.9 (+0.3 to +1.4) | +1.0 (+0.4 to +1.6) |
| Serious | No majority | +0.1 (-0.3 to +0.5) | +0.1 (-0.3 to +0.5) | -0.1 (-0.4 to +0.3) |
| Non-critical | Majority-Black | +2.7 (+1.3 to +4.3) | +2.6 (+1.1 to +4.4) | +1.2 (+0.0 to +2.6) |
| Non-critical | Majority-Hispanic | +3.0 (+1.5 to +4.4) | +4.9 (+3.0 to +6.9) | +7.0 (+4.1 to +10.0) |
| Non-critical | No majority | +0.4 (-0.2 to +1.0) | +0.5 (-0.3 to +1.3) | +0.7 (-0.2 to +1.6) |
| Not a crime in progress | Majority-Black | +10.4 (+6.3 to +15.0) | +8.4 (+4.1 to +14.4) | +5.4 (+1.3 to +10.0) |
| Not a crime in progress | Majority-Hispanic | +14.9 (+10.7 to +19.6) | +18.6 (+13.4 to +24.4) | +19.3 (+13.7 to +26.0) |
| Not a crime in progress | No majority | +1.9 (-1.0 to +4.8) | +1.8 (-1.1 to +4.9) | +1.7 (-1.5 to +4.7) |
The data shows differences in recorded times between groups of neighborhoods. It does not show why: how units were deployed, what else they were handling, or how dispatchers chose.
Times depend on which calls are made, from where, and how call takers classify them. Neighborhoods differ in all three; holding the call type fixed does not hold fixed the judgment behind it.
The clock starts when the call taker enters the job, not when the phone rings, and stops when a unit is marked as arrived. Officers mark their own arrival, and some calls never get an arrival time.
Groups describe neighborhoods by their residents' income and makeup (Census estimates). The data says nothing about who made a call.
The data has no field for whether an event started with a call. 15,560,037 events were left out: directed patrols and visits (75 codes), officers asking for assistance (13 codes), ShotSpotter alerts, and events with an arrival logged in the same second as the entry or dispatch.
Calls with no recorded arrival are left out of the waits. For calls not about a crime in progress their share is highest in both the lowest-income fifth (45%) and the highest-income fifth (42%), and in majority-Hispanic (45%) and majority-White (43%) neighborhoods, against 37% in majority-Black neighborhoods. Counting them as waiting until the event was closed widens the gaps for less urgent calls.
Officers per call counts uniformed officers assigned to each precinct on 2026-10-04, after the window, not officers on patrol in a given hour. It is the only public precinct staffing figure found.
Distance runs from the call to its precinct's station house. Patrol cars are usually out on patrol, not at the station.
Only 6 neighborhoods are majority-Asian. Their results are in the tables but not in the headline.
The 116th Precinct was formed from part of the 105th in December 2024. Each call keeps the precinct it was coded to at the time.
Calls from NYPD's dispatch data on NYC Open Data, 2022 to 2025, with January to June 2026 held out as a check. Rows that repeat an event are combined, keeping the first entry, dispatch and arrival. The wait runs from the call being entered (ADD_TS) to the first arrival (ARRIVD_TS), for calls with an arrival within 24 hours; each of NYPD's four priority categories is analyzed on its own. Neighborhoods are the 197 residential 2020 Neighborhood Tabulation Areas, grouped by median household income (fifths of residents) and by the group that is more than half of residents (Hispanic of any race; Black, White and Asian non-Hispanic), from the ACS 2024 5-year estimates. Each group's calls are reweighted to the citywide mix of call type, hour of the week and precinct busyness; 95% intervals come from 2,000 resamples of neighborhoods, and p-values are Holm-adjusted across the 20 planned comparisons. The plan, with the smallest gaps that matter (1 minute for Critical calls, 20% for the rest), was committed before the analysis ran. Citywide medians match the city's 911 End-to-End figures.
NYPD Calls for Service, Historic (d6zx-ckhd) and Year to Date (n2zq-pubd); NYC 911 End-to-End Data (t7p9-n9dy); 2020 Neighborhood Tabulation Areas (9nt8-h7nd), 2020 Census Tracts (63ge-mke6) and Police Precincts (y76i-bdw7), all NYC Open Data; NYPD precinct list, geocoded with NYC Planning GeoSearch; NYPD Officer Profile (pmsy-ewrc), counts by command; American Community Survey via Census Reporter.