How Traffic Studies Actually Get Done — and Who Pays for Them

You asked the city for a stop sign and got "it didn't meet the warrant." Here's what that sentence actually means, who counted the cars, and who paid.

You stand at the corner where the car almost hit your kid, and you call the city. A few weeks later a letter comes back: the location "did not meet the warrant for a stop sign." You read it twice. Nobody disagrees that drivers fly through. So what is a "warrant," who decided, and why does the obvious thing not happen?

That gap — between what a neighbor feels and what a traffic study finds — is where most of the frustration lives. The good news is that almost none of this is secret. It runs on counted numbers, published thresholds, and meeting agendas you can read. Here's how the machine actually works.

Step one: somebody counts the cars

A traffic study starts with raw counting, and the counting is more boring and more rigorous than people expect. The classic tool is the pneumatic road tube — the black rubber hose stretched across the pavement that clicks every time an axle rolls over it. Tubes measure volume (how many vehicles), and paired tubes measure speed and vehicle length, which is how a city estimates the share of trucks.

For turning movements and pedestrians, cities use human counters with clickers or, increasingly, video and sensor counts. The standard outputs you'll hear quoted:

• ADT / AADT — average daily traffic, the typical number of vehicles per day.
• Peak hour volume — the busiest 60 minutes, usually the morning and evening commute, because intersections are designed around their worst hour, not their average.
• The 85th percentile speed — the speed at or below which 85 out of 100 drivers travel. This single number drives a huge amount of speed-limit and enforcement debate, and it surprises people: limits are often set near what most drivers already do, not below it.

Counts are usually taken on mid-week, non-holiday days in normal weather, because one rainy Monday or a school break can skew the picture. If you ever request a study, check the dates — a count taken during summer recess near a school is a fair thing to question.

Step two: the "warrant"

Here's the word that ends so many requests. A warrant is a published threshold that has to be met before a particular traffic control is considered justified. The national reference is the Manual on Uniform Traffic Control Devices (the MUTCD), maintained federally, which most states adopt or adapt. It lays out warrants for things like traffic signals, multi-way stop signs, and school-zone treatments.

Signal warrants, for example, look at things like eight-hour and peak-hour volumes, pedestrian crossing volume, crash history, and proximity to other signals. A multi-way stop has its own criteria around volume balance between the streets and collision counts. The core idea is consistency: a driver should be able to trust that a stop sign or signal means roughly the same conditions everywhere, and that signs aren't sprinkled around on vibes.

The honest tension: warrants are designed to prevent over-signing, but they can feel cold when a place is dangerous in a way the thresholds don't capture — a blind curve, a street kids cross to a park, near-misses that never became a logged crash. Warrants are a floor for justification, not a ceiling. Engineers can use judgment, and elected officials can direct money. "It didn't meet the warrant" often really means "it didn't auto-qualify, and nobody pushed it further."

Who pays

Roughly three buckets:

• The public agency. For existing roads, the city, county, or state DOT typically funds the study out of its traffic-engineering or public-works budget. This is what happens when you, the resident, file a request.
• The developer. When someone proposes a new subdivision, store, or apartment complex, the city usually requires a traffic impact study paid for by the developer as a condition of approval. The developer hires a private traffic-engineering firm, and that consultant's report estimates new trips and recommends mitigations — a turn lane, a signal, a wider approach.
• Grants. Safety-specific pots of money (federal and state) can fund studies and fixes, especially after a corridor shows a crash pattern.

That developer-pays model is where the politics get real. The applicant pays the consultant whose study can make or break the project. Reputable firms protect their licenses and their data, and cities can require independent peer review. But it's exactly why you should read who commissioned a study before you trust its conclusions — not to assume bad faith, but because trip-generation assumptions and the choice of which intersections to even include are judgment calls with real room to lean.

How to actually use this

If you want something changed, work the process the way it's built:

1. Ask for the study, and the raw counts. Most traffic data is a public record. Request the count sheets, the dates, and which warrant was checked. "Show me the numbers" is a fair, normal ask.
2. Document the part the count missed. Photos, dates, and times of near-misses; the school bell schedule; where kids actually cross. Crash-history gaps are common because minor collisions go unreported — your log can fill that in.
3. Bring neighbors and go political. Warrants gate the engineering recommendation, but a city council or county board can fund a fix anyway. A petition and three minutes at public comment move things that an email never will.
4. Watch developer studies near you. When a project posts its traffic impact study, you can comment on its assumptions during the approval window — and ask the city for independent review.

The system isn't rigged so much as it's rule-bound and quiet. The counts are real, the thresholds are published, and the meetings are open. The neighbors who get results are the ones who stop waiting for the letter and start reading the same records the engineers do.