What 'Complete Streets' Policies Actually Require
Jump ahead
A council passes a complete streets resolution. There's a press release, a photo of seven people holding a printed copy, a line in the local paper about the city's new commitment to safe streets for all users. Five years later, the four-lane arterial by the high school still has a sidewalk on one side that dead-ends at a utility pole, and the crossing is still eighty-odd feet of unmarked asphalt.
Nothing malfunctioned. The resolution did exactly what resolutions do.
If you want to know whether your city's policy will change anything you can walk on, stop reading the vision statement. Go find four things: what triggers the policy, who can waive it, which design manual it points to, and whether anyone has to report results in public. The rest is framing.
The document is not the street
"Complete streets" is a policy approach, not a street type. The core idea is boring and hard to argue with: when a public agency designs, builds, or rebuilds a street, it should consider everyone who needs to use that street — people walking, people using wheelchairs, people on bikes and scooters, people waiting for a bus, freight, and yes, drivers — rather than defaulting to a design whose only measured objective is moving cars through quickly.
That's it. That's the whole concept. Everything contentious lives in the implementation.
Which is why a policy can be genuinely well-written and still produce nothing. Street design in most places is governed by a stack of engineering standards, state DOT control over certain roadways, utility coordination, funding eligibility rules, and a public works department with a paving schedule set years in advance. A resolution sits on top of that stack without touching any of it. To move the street, the policy has to reach down into at least one of those layers and change a default.
Resolution, ordinance, or design manual?
This is the single most useful distinction, and almost nobody in a public meeting makes it. "The city has a complete streets policy" can mean four wildly different things.
| Instrument | How binding | What it usually contains | What it can't do |
|---|---|---|---|
| Resolution | Statement of intent; not enforceable | Vision language, list of user groups, direction to staff to "consider" or "develop guidance" | Compel any specific project change |
| Ordinance / municipal code | Law; enforceable, survives turnover | Applicability triggers, exception process, who signs off, reporting requirements | Specify lane widths or crossing spacing in useful detail |
| Executive / departmental directive | Binding on staff while the administration holds | Internal review checkpoints, project checklists, sign-off routing | Bind a future administration |
| Design manual or standard details | Binding on drawings; the operative document | Actual dimensions, cross-section menus, intersection treatments, standard detail sheets | Force a project to get funded or scheduled |
A city with a beautifully written resolution and a 1970s design manual will keep building 1970s streets. A city with no complete streets language at all but a modern design manual and an updated set of standard details will quietly produce better streets than its neighbors. I'd take the second city every time.
The strongest arrangement pairs an ordinance (which creates the obligation and the paper trail) with an updated manual (which tells engineers what to draw). One without the other is theater or luck.
What actually triggers the policy
A policy that applies to "all transportation projects" sounds maximal and is often weaker in practice than one that names specific triggers, because "all projects" invites argument about what counts as a project. Resurfacing? Utility cut restoration? A developer's frontage improvement? A signal rebuild?
Resurfacing is the big one, and it's where a lot of the real opportunity sits. When a street gets milled and repaved, the striping comes off. Repainting it differently — narrower travel lanes, a bike lane, a wider parking-side buffer, tightened corner radii — costs a fraction of a reconstruction because the paving crew is already there and the paint is going down anyway. Cities that write resurfacing into their trigger list get a slow drip of cheap improvements. Cities that exclude it wait for full reconstructions that come around every few decades per street.
| Project type | Typical cost of adding people-first elements | Commonly in scope? |
|---|---|---|
| Full reconstruction (curb to curb, new base) | Low as a share of a very large budget | Almost always |
| Mill and overlay / resurfacing | Very low if it's restriping only; high if curbs move | Varies the most — check this clause first |
| Signal or intersection rebuild | Moderate — crossing geometry and timing are already open | Often, sometimes only for the signal itself |
| Bridge deck replacement | Moderate to high; width is fixed by structure | Frequently exempted on structural grounds |
| Private development frontage | Borne by developer, not the city | Usually handled in zoning code instead |
| Utility trench restoration | Low, but scattered and hard to coordinate | Rarely |
One more trigger worth hunting for: does the policy apply to projects on state-owned roads inside city limits? Many of the worst streets in any American city — the wide, fast, retail-lined stroads — are technically state highways. A municipal policy generally can't rewrite a state DOT's standards. What it can do is establish the city's formal position in negotiations, which matters more than it sounds when the DOT needs local buy-in.
Exceptions are where these policies go to die
Every serious complete streets policy has an exception process. It has to. There are real cases: a limited-access freeway ramp where pedestrian facilities would be genuinely unsafe, a segment where the right-of-way is physically 22 feet wide and bounded by 1890s buildings, a project so small that mobilizing for anything extra costs more than the benefit.
The question isn't whether exceptions exist. It's how expensive they are to use.
A weak exception clause lets a project engineer note "not feasible due to constraints" in a field on a form and move on. Nobody outside the department ever sees it. Over a decade this produces a policy with a 90 percent exception rate and a city that sincerely believes it has complete streets.
What a costly exception looks like
- Named signatory. Not "the department" — a specific title, ideally the public works director or city engineer, personally signing each waiver.
- Written justification against listed criteria. The policy enumerates the acceptable reasons. Anything outside the list isn't an exception, it's a policy violation.
- Documented alternatives. If the full treatment doesn't fit, what interim or partial measure was considered, and why was that rejected too? "No room for a protected bike lane" doesn't explain skipping a crosswalk.
- Public posting. Waivers published on a webpage or attached to a periodic report. Sunlight does more here than any amount of vision language.
- Cost is a reason, not a magic word. Good policies require the cost claim to be quantified as a share of total project cost. "Too expensive" collapses fast when the number is 1.8 percent.
If a policy has no reporting requirement, assume you will never find out how often it's being waived — and neither will the council members who voted for it.
The other silent killer is the "no available right-of-way" claim. Sometimes true. Often it means "no room without removing a travel lane or on-street parking," which is a choice about allocation, not a physical constraint. A good policy forces that distinction onto paper: state the curb-to-curb width, state the current allocation, then state why the current allocation is fixed.
The design manual does the real work
Ask what your city's default travel lane width is on a 30 mph collector. If the answer is 12 feet, the policy hasn't reached the manual yet.
Twelve-foot lanes are a highway dimension that migrated into cities and stayed. On a low-speed urban street, narrower lanes give you back space for everything else and don't hurt safety at city speeds — and the reclaimed feet are exactly what a bike lane or a wider sidewalk needs. Lane width is the clearest single tell for whether a design manual has actually been modernized.
A few other things worth checking in the manual, because they show up in almost every project and almost never in the press release:
- Corner radius defaults. Big sweeping curb returns let drivers turn fast and lengthen every crossing. This is a cheap fix on any signal or curb project.
- Maximum crossing spacing. If the manual doesn't cap the distance between marked crossings, you get quarter-mile gaps and people crossing mid-block anyway, then getting blamed for it.
- Standard detail sheets. The drawings contractors actually build from. If there's no standard detail for a concrete-protected bike lane or a bus boarding island, every instance becomes a custom design, which means it gets value-engineered out.
- Whether the manual references current national guidance. A manual whose most recent citation is decades old will keep producing decades-old streets no matter what the ordinance says.
Follow the money
Complete streets language often enters a city sideways, through funding eligibility. Regional and federal grant programs increasingly ask applicants to describe their multimodal policies, and some states condition certain transportation funds on having one. That's a real lever, and it also explains a lot of thin policies: if a resolution unlocks grant scoring points, a resolution is the rational minimum to adopt.
Which means the honest read on some cities is that the policy exists to win money, not to change streets. That's not nothing — the money builds things — but don't confuse the two.
Complete doesn't mean identical
The most common misreading, and it comes from both directions.
Opponents hear "complete streets" and picture a mandate to bolt a bike lane onto every road in the city, including the industrial spur with four trucks a day and the residential lane where a car passes every three minutes. Advocates sometimes argue as if that's the goal.
It isn't, and the better policies say so explicitly. A complete street is one where the design matches the network role and the surrounding context. On a quiet residential street with low volumes and low speeds, the complete solution can be a shared roadway with nothing added — because the mixing already works. On a 40 mph arterial with a bus route, big-box frontage, and a school, separated facilities aren't a nicety; without them the street is effectively closed to anyone not in a vehicle.
That's why modern policies pair with a street typology: a classification that sorts streets by the kind of place they run through and the kind of movement they carry, then attaches a design menu to each type. Without a typology, "complete" gets litigated street by street, and the loudest room wins.
The counterintuitive part: the strongest complete streets policies allow more variation across the network, not less. They just remove the ability to choose "nothing" by default.
Reading your city's policy in twenty minutes
Search your city's site for "complete streets" plus "ordinance" or "chapter." If you only find a resolution, search the public works page for a design manual or engineering standards. Then, in order:
- Find the applicability section. Does it list resurfacing? Does it cover state routes within city limits? Does it cover private development frontage?
- Find the exceptions section. Who signs? Is the justification written? Is it published anywhere?
- Find the reporting requirement. Annual report to council, with project counts and waiver counts, or nothing? If nothing, the policy is unmeasured by construction.
- Find the design reference. Does the policy name a manual, and does that manual have a revision date in the current decade?
- Check the default lane width and corner radius. Two numbers, and they'll tell you more than the whole preamble.
If items 3 and 4 are missing, the useful thing to ask for at the next meeting isn't stronger vision language. It's a reporting requirement and a manual update — the two least quotable asks available, and the two that actually move concrete.
FAQ
Does a complete streets policy mean the city has to remove parking?
No. Nothing in the concept requires it. Removing parking is one option for finding space on a constrained street, and it's the option that generates the most conflict, so it dominates the conversation. Narrower lanes, a reduced number of travel lanes, or a widened sidewalk on one side only are all more common outcomes.
Do these policies apply to highways?
Generally not to limited-access highways, and legitimately so. The tricky category is the state-owned surface arterial running through a downtown, which is functionally a city street with highway ownership. Rules there depend on your state DOT and on any agreements it has with the city.
How long until you'd expect to see changes?
On a resurfacing-triggered policy, changes appear on the current paving cycle — sometimes within a year on whatever streets were already scheduled. On a reconstruction-only policy, decades, because that's how long it takes to cycle through a street network. This is the practical reason the resurfacing clause matters so much.
What's the difference between complete streets and Vision Zero?
Different starting points. Complete streets is about who a street is designed for; Vision Zero is a commitment to eliminating traffic deaths and serious injuries, usually with a high-injury network analysis driving where you spend first. They overlap heavily in practice and cities often adopt both, but a Vision Zero plan typically comes with data and a prioritized project list, which makes it easier to hold anyone to.
Can a complete streets policy be repealed?
Yes, and a resolution or a departmental directive can simply be ignored or quietly rescinded after an election. Code changes and design manual revisions are far stickier, because reversing them means an affirmative vote or a formal engineering process someone has to own. Durability is a design choice too.