Research · 4 min read
Seattle’s Slowest 911 Waits Are Queues, Not Drives
Response time is a distribution dominated by dispatch delay for lower priorities, not a single average.
Response time is usually sold as a single average. Seattle’s public call data says the useful number is a distribution — and that most of the wait for lower-priority calls is queueing, not driving. Priority-one calls hit seven minutes about half the time. Priority-two calls rarely do, and priority-three actually beats them.
Our analysis of 250,000 sampled calls, January 1–December 31, 2024.
Key takeaways
- Priority-1 calls had a 6.7-minute median and met a seven-minute target 52.5% of the time; priority-2 had a 21.1-minute median and met it only 22.5% of the time.
- Priority-3 beats priority-2 on that same target (35.7% vs 22.5%), so priority signals intent, not speed.
- Most of the wait for lower priorities is dispatch delay, not travel: 14.8 minutes of the priority-2 median and 29.7 of priority-3.
- About 8.8% of sampled calls had no recorded response time, and the causal-evaluation harness is a documented design template, not a finding.
What does Seattle’s public call data reveal about response-time priority?
Across 250,000 sampled Seattle calls from 2024, priority-1 calls had a 6.7-minute median and 52.5% arrived within seven minutes; priority-2 had a 21.1-minute median of which 14.8 was dispatch delay, yet priority-3 beat it on the seven-minute target (35.7% vs 22.5%). About 8.8% of calls had no response time. Priority signals intent, not speed.
The numbers behind the answer
Selected measures only. Denominators and interpretation stay attached so the headline cannot stand alone.
6.7 min
Priority-1 median response
Across 62,638 priority-1 calls; 52.5% arrived within seven minutes.
14.8 min
Priority-2 median dispatch delay
Of a 21.1-minute median priority-2 response, most is dispatch delay.
8.8%
Calls with no recorded response
21,949 of 250,000 sampled calls lack a usable response time and drop out of percentiles.
Read the middle and the tail, not the mean
Across a 250,000-call sample, priority-1 calls have a median response of 6.7 minutes, but the 90th percentile is 16.6 and the 95th is 22.0. Priority-2 calls sit at a 21.1-minute median with a 185-minute 90th percentile, and priority-3 at a 31.0-minute median with a 349-minute 90th. A single average would erase all of that structure.
Roughly eight percent of sampled calls carry no response time at all — 21,949 of 250,000, or 8.8%. Any scorecard that reports only an average is reporting neither the tail nor the missing rows.
The percentiles also move in a way a mean cannot. Priority-1 has a tight 16.6-minute 90th percentile; priority-2 and priority-3 stretch past three and nearly six hours at the same point. That gap between the typical call and the slow tail is the whole operational story, and it is invisible in a blended number.
Evidence visual
Median response time by call priority
Minutes from call queued to first unit arrival, 2024 sample.
Most of the wait is dispatch delay
Response time here runs from the call being queued to the first unit arriving, which includes time spent waiting to be dispatched and time spent driving. Splitting them changes the story: priority-1 calls carry a 1.48-minute median dispatch delay, but priority-2 carries 14.78, priority-3 29.73, and priority-5 45.2. The delay, not the drive, is where the time goes.
For lower priorities, the problem is largely queueing, not distance. That distinction matters for any capacity or staffing question — one points at how many units are available and how calls are triaged, the other at geography and traffic. A blended average cannot tell you which, and it is the first thing to decompose before drawing a conclusion.
Scorecard rule
Separate dispatch delay from travel time before drawing any staffing or distance conclusion.
Priority-3 beats priority-2 on the seven-minute target
The clearest surprise in the data is that priority tiers are not ordered by speed. Priority-3 calls have a higher median than priority-2 (31.0 vs 21.1 minutes) yet meet the seven-minute target more often: 35.7% versus 22.5%. Priority-4 goes further, hitting seven minutes 37.0% of the time with a 43.5-minute median. Priority-1, the only tier that behaves as expected, clears seven minutes 52.5% of the time.
The likely explanation is triage and queueing. Priority-2 is a broad middle tier that absorbs many calls, and its 14.8-minute dispatch delay is where the target is lost; priority-3 and priority-4 include quick, easily closed incidents. Whatever the mechanism, the scorecard lesson is fixed: priority is a statement of intent, not a measurement of how fast a call will actually be answered.
A reusable evaluation harness for policy questions
The program also publishes a reusable causal-evaluation harness for policy questions — difference-in-differences and event studies on a beat-week panel. It is a documented, importable design that can be applied to future policy questions.
It is deliberately unfinished as a result. The harness is a pre-registered design template that cannot produce a verdict until an intervention-date registry exists; it reports itself unavailable rather than manufacturing an effect. That restraint is the point — a template is only useful if it is honest about what it cannot yet say.
Nothing here describes 911 Sentinel’s responder model, and none of it is a service-level promise. It is what a public record of police calls actually shows.
- Report a target-hit rate and upper percentiles, not just a mean.
- Separate dispatch delay from travel time.
- Show how many calls have no recorded response.
- Exclude autogenerated traffic and admin priorities.
- Pre-register any causal design before looking at outcomes.
The slow minutes sit in the low priorities, and dispatch is half the wait
The tail is concentrated exactly where it is least defended. Priority-1 calls had a 90th percentile of 16.55 minutes and a 95th of 21.97; priority-3 calls reached 349.12 and 537.26. The median response runs from 6.72 minutes at priority 1 to 44.87 at priority 5, and the median dispatch delay ranges from 1.48 minutes at priority 1 to 45.2 at priority 5 — so at the lowest priority, well over half the clock passes before a unit is even assigned.
Two numbers keep the distribution honest. 21,949 of 250,000 sampled calls (8.78%) had no recorded arrival and are censored, which biases any naive median downward; and the share answered within seven minutes swings from 22.53% at priority 2 to 52.52% at priority 1. A single median hides both the tail and the missing rows.
Why it matters is an outcome question. Quasi-experimental work links faster police response to higher clearance, and in cardiac arrest shorter ambulance response raises survival. The tail is where the consequence lives, so the defensible target is a wait the property’s own risk can justify — not an average.
Why it matters
Response time is worth measuring because it changes clearance and outcomes, not only satisfaction.
Questions property teams ask
Does this replace a response-time average?
It complements it. A median, a target-hit rate, and upper percentiles describe the distribution that a single mean hides.
Why separate dispatch delay from travel?
They imply different fixes. For lower priorities, most of the wait is time before a unit is dispatched, not driving distance.
Why does priority-3 beat priority-2 on the seven-minute target?
Priority tiers are not ordered by observed speed. Priority-2 absorbs many calls and carries a 14.8-minute dispatch delay, while priority-3 includes quick, easily closed incidents.
What is the difference between queued time and response time?
Queued time is when the call enters the system; response time runs from queue to first unit arrival and includes both dispatch delay and travel.
What are the missing 8.8%?
Calls with no usable recorded response time. They are counted and disclosed rather than silently dropped.
What does the evaluation harness do?
It packages a difference-in-differences and event-study design on a beat-week panel so the same public call data can be reused to evaluate future policy questions. It is a template, not a result.
Does response time actually change outcomes?
Yes, in measurable ways. Research links faster police arrival to higher clearance rates, and faster ambulance arrival to better survival in cardiac arrest. The distribution is both a service metric and an outcome variable.
Our methods, limits, and sources
How we calculated this
This is original 911 Sentinel research — we gathered the records, ran every calculation below, and published the aggregate dataset.
We sampled 250,000 Seattle CAD calls from 2024 across priorities 1–6 and computed response-time distributions, dispatch-delay decomposition, target-hit shares, and censoring.
- We excluded priorities 7, 9, and blank administrative categories.
- We converted response and dispatch-delay seconds to minutes and computed percentiles by priority.
- We decomposed response into dispatch delay and travel.
- We computed the share of each priority arriving within seven minutes and recorded censoring.
- We documented a reusable causal harness and marked it unavailable absent an intervention-date registry.
What this analysis cannot establish
- Response time is queued-to-first-arrival and includes dispatch delay; decompose before comparing.
- Location is beat-blurred with no unit-level geography.
- About 8.8% of sampled calls lack a response time and are excluded from percentiles.
- Priority tiers are not ordered by observed speed, so they cannot be treated as a service ranking.
- Autogenerated traffic and administrative priorities are excluded.
- This is public CAD data, not the internal responder model, and is not a service-level promise.
Sources
The raw records come from the sources below; the study design, analysis, charts, and conclusions are our own.
- Seattle Police Department — Call Data
- Published aggregate result (JSON)
- Source manifest (JSON)
- Jordi Blanes i Vidal & Tom Kirchmaier (The Review of Economic Studies) — The Effect of Police Response Time on Crime Clearance Rates
- Johan Holmén et al. (Journal of the American Heart Association) — Shortening Ambulance Response Time Increases Survival in Out-of-Hospital Cardiac Arrest
Related questions and practical guides
No obligation · free property walk
Ask for the distribution, not the average
Any monitoring or response proposal should show a target-hit rate, upper percentiles, and how missing calls are handled — not a single blended number.