The NCUTCD Wants to Know How You Think Speed Limits Should Be Set

A neighborhood greenway in Portland, Oregon. Photo: public domain.
A neighborhood greenway in Portland, Oregon. Photo: public domain.

PlacesForBikes is a PeopleForBikes program to help U.S. communities build better biking, faster. You can follow them on LinkedIn, Twitter or Facebook or sign up for their weekly news digest about building all-ages biking networks.

This doesn’t happen very often: The obscure but hugely influential National Committee on Uniform Traffic Control Devices is asking for direct feedback from ordinary transportation pros.

The NCUTCD is the body of transportation engineers that largely controls the shape, design, and content of all U.S. traffic signs and signals, as well as various features of roadway design including the posted speed limit.

Now, informed by a groundbreaking 2017 report about the importance of traffic speed to traffic safety, the NCUTCD is considering recommending changes in how local governments should set speed limits: not based only on how fast people naturally tend to drive on a stretch of road, but also on how the public wants people to drive on a stretch of road.

The leader of the NCUTCD task force posted a 13-question online survey Wednesday asking for input. It takes about 5 minutes to complete.

For decades, it’s been an article of faith among many traffic engineers that posted speed limits should be set at approximately the 85th percentile of observed auto speed — that is, the speed at which 15 percent of motor vehicles are driving faster and 85 percent are driving slower. That principle was enshrined in one of the traffic engineers’ bibles: the Manual on Uniform Traffic Control Devices, maintained by the Federal Highway Administration and largely written by the NCUTCD.

But some NCUTCD members like Dongho Chang, Seattle’s city traffic engineer, argue that the 85th percentile principle has been overinterpreted, contributing to unnecessarily dangerous streets.

“There’s a lot of misconceptions from practitioners that it should be set at 85th percentile, which is really not the intent of how the MUTCD is supposed to be applied,” Chang said Wednesday. “You’re supposed to use context for setting speed limits.”

Even inside a car, speeding sharply increases the risk of serious or fatal injury. Image: National Transportation Safety Board.

Last summer, when the National Transportation Safety Board recommended backing off from the 85th percentile principle, we called auto speed the invisible X factor in bicycle infrastructure. The difference between a 22 mph car and a 32 mph car isn’t just the difference between life and death if something goes wrong — it’s also the difference between pleasant and harrowing in the best-case scenario.

That’s why the posted speed limit on each street is one of the crucial factors used in PlacesForBikes’s own Bike Network Analysis of 500 U.S. cities.

Because nobody likes to spend their days contemplating their possible demise, auto speed is also the difference between bikes being useful and relevant tools for most Americans’ daily life or being toys most Americans use once in a while.

The NCUTCD survey is very technical in nature, and if you don’t have a background in traffic engineering it’s really not for you. But if you’re a working professional in the field — or if you know someone who is — it’s definitely worth knowing about.

36 thoughts on The NCUTCD Wants to Know How You Think Speed Limits Should Be Set

  1. … not based only on how fast people naturally tend to drive on a stretch of road, but also on how the public wants people to drive on a stretch of road.

    Then we really should focus on not designing roads to be driven on fast, not just rushing to put up new signs.

  2. One issue with the survey is that it is biased toward high-speed roads, and not urban / town center / pedestrian priority roads. The current 85th percentile / USLIMITS2 system doesn’t leave much room for intentionally designing for slower speeds in those types of areas to improve safety for all road users (while improving the local economy with walkable / bike-able streets). I left several comments to that effect to counteract (and point out) the survey bias toward car-centric speed and flow.

  3. On a stretch of road near me they lowered the speed limit from 30 to 25. Driving down the street at the new speed limit, I got overtaken several times. Did lowering the speed limit really accomplish much?

  4. I go out of my way to engage traffic engineers. They’re an interesting group, not at all like the average Streetsblog reader. Their mission is to move as much traffic (without regard to what it looks like, though it obviously is mostly cars, trucks, buses, etc.) as quickly, safely and cheaply as possible. If lowering speeds meshes with that, they’ll support it. If it doesn’t, they don’t really give a rip about all the social ramifications.

  5. In the absence of enforcement close to 24/7/365 AND frequently enough along the street (maybe every 1.5 to 2 blocks) which are levels of enforcement that no city can afford, posted speed limits have almost no effect on the actual 85th percentile speeds (plus or minus 0 to 3 mph). The only way to change an 85th percentile speed of about 35 mph to one of 25 is to re-engineer the street so the slowest 85% of the drivers now feel safe and comfortable only at speed up to about 25. This is effective, but it is expensive and may have some negative effects for congestion, economic factors if some shoppers stop coming, and for diversion of some vehicles onto smaller roughly-parallel streets making them significantly more hazardous as frustrated drivers attempt to go around the “newly-installed” congestion. Re-engineering also has another serious negative for some cities, it is then no longer to profiteer with part-time enforcement that does not lower travel speeds but does make big profits. Engineers DO have to try to move vehicles efficiently and safely because our economy depends upon mobility for commuters, shoppers, tourists, and commercial traffic – into and out of our cities and commercial areas.

    James C. Walker, National Motorists Association

  6. It is EASY to design slower speeds by designing the roads that way. But it may cause issues of congestion, diversion of traffic, and reduction of some shoppers that will then refuse to accept the congestion.

    James C. Walker, National Motorists Association

  7. How about the US road fatalities at 4x rate of the EU? Seems like we’re doing it wrong, systemically.

  8. For starters, it is not 4x. Fatality rates per billion km traveled: US-7.1, UK-3.6, France-5.8, Germany-4.9, Austria-5.8, Belgium-7.3. Then look at the vast distances in the USA, the relative lack of good transit in so many areas, and the much higher fatality rate per distance in very rural areas where times for EMS and transport to a decent trauma hospital are so much longer. Easy car ownership started early in the 20th century in the USA and was 30-50 years later in Europe – factors that had huge effects on how the mobility systems developed. It just is not an apples to apples comparison. We COULD redesign some inner cities for more town center/pedestrian priority areas. But you are talking multi-billion dollar changes including serious improvements to parking at reasonable prices, so shoppers arriving from the suburbs and beyond don’t just abandon what is left of the inner city shopping areas as just too much trouble to be worth the hassles.
    James C. Walker, National Motorists Association

  9. James – Ahem. You are selectively quoting the rate of fatalities per mile traveled. A more useful rate is per 100K population. Trying to make the fatality rate look less bad because US development has resulted in unsustainable (deadly) sprawl is sneaky, and also props up sprawl as the best future state. So yeah. Your statements are intentional propaganda for an unsustainable and dangerous system.

  10. Yes, please do examine our website. The “difference” in our view is that we approach issues from the standpoint of 75+ years of traffic safety engineering for what works, and what does not work. Artificially low posted limits and part time enforcement are about $$$$, not safety.

    James C. Walker, National Motorists Association

  11. Please note that in a great many states it is not urban sprawl – it is the rural character of the areas. Where people have to drive 20+ miles to their nearest reasonable city for services, the rate of events per mile traveled is the right criteria.

    James C. Walker, National Motorists Association

  12. I would call into question the loss of shoppers to businesses as a calmer street will likely lead to a more pleasant shopping experience, but otherwise agree completely. Putting up signs with lower numbers is nonsense. Either redesign the street or at least put up speed cameras. Anything else is just a continued dereliction of duty.

  13. Just because someone has horses doesn’t mean that they live in a “rural” area. Most people live in an either urban area or area of urban sprawl. With their horses.

  14. WE like data as well, and base most of our principles on data. But miles traveled is a key factor. If you have X number of fatalities with Y numbers of miles driven – then X stays about the same but Y becomes about 10% more, then the fatality rate went down by about 10%.

    James C. Walker, National Motorists Association

  15. You make a good point for the shoppers who live very nearby or who want to arrive by transit – fewer vehicles can make the experience better. But if the area depends upon a lot of shoppers who want to arrive by car and on employees who need to arrive by car because of where they live – that changes the potential equation.

    James C. Walker, National Motorists Association

  16. In some areas this is true, but in many areas it is not true. We have both in the US.

    James C. Walker, National Motorists Association

  17. It’s true for the vast majority of Americans. The lion’s share of the population lives in an area that would not be best described as “rural” by all but a big stretch of the word.

  18. Probably true in total numbers, but many “urban” areas are almost totally car necessary with inadequate transit options. This is particularly true where the areas have some sprawl and it takes 2X or 3X or more time to get to many destinations than the 1X time it takes to drive. Those areas also contribute seriously to the total miles driven – making events per VMT a more accurate comparison parameter than events per population. We drive over 3 trillion miles per year with a bit over 200 million drivers.

    James C. Walker, National Motorists Association

  19. Slowing traffic to the average speed of bikes is the most-critical issue for cycling? If 3% of all traffic is cyclists, why should the other 97% take a huge hit to transport speed, which is a form of productivity, especially for commercial traffic?

    Perhaps it would be much-safer for cyclists to not attempt to mix it up with heavy high-speed traffic then and to stay on more lightly-traveled slower-speed routes?

  20. The rate per-distance traveled is in-fact the same rate that the ICC used to use, and is more-representative of the majority of US road travel, which averages a much longer distance per-person than Western European road travel, which like NY or NJ travel involves a much higher percentage of public transit use than is common in other States.

    Fatalities/100,000 population:

    Arizona: 18.77
    California: 9.37
    Colorado: 10.92
    New Mexico: 19.3
    Texas: 13.59
    Utah: 9.18
    Wyoming: 25.98

    By your fatalities per-population statistic it looks like Wyoming has terrible highway safety, however, what happens if we change how we measure traffic safety to a statistic that measures the fatality rate per square mile?

    Fatality Rate per Square Mile
    State —– Square Area — Deaths 2016 —- Deaths/Sq Mi
    Arizona —— 113,998 ———– 962 ————- 0.0084
    California —-163,696 ——— 3,680 ————- 0.0225
    Colorado —- 104,185 ———– 605 ————- 0.0058
    Nevada ——-110,567 ———– 327 ————- 0.0030
    New Mexico – 121,697 ————404 ————- 0.0033
    Utah ———– 84,899 ———— 280 ————- 0.0033
    Texas ———268,597 ——— 3,773 ————- 0.0140
    Wyoming —– 97,093 ————-152 ————- 0.0016

    Connecticut —-5,543 ————- 311 ————- 0.0561
    New Jersey —- 8,723 ————- 607————– 0.0696
    New York —– 54,556 ———– 1,025 ————- 0.0188
    New York City — 304.6 ———— 229 ———— 0.7518
    New York State
    less NYC —– 54,251 ————– 796 ————- 0.0147

    NYC ped. fatalities 2016: 144, or 62.9% of traffic fatalities
    NYC bike fatalities 2016: 18, or 7.9% of traffic fatalities
    NYC driver fatalities 2016: 67, or 29.8% of traffic fatalities

    CO ped. fatalities 2016: 84 or 13.9% of traffic fatalities
    CO bike fatalities 2016: 16, or 2.6% of traffic fatalities

    NV ped. fatalities 2016: 57, or 17.4% of traffic fatalities
    NV bike fatalities 2016: 9, or 2.8% of traffic fatalities

    NM ped. fatalities 2016: 77, or 19.1% of traffic fatalities
    NM bike fatalities 2016: 8, or 2.0% of traffic fatalities.

    Looks like the most-dangerous place to drive, walk, or ride a bike by far is New York City, which has a fatality rate per-square mile 470 times as bad as that of Wyoming!!!

    [Quote] “New York State still holds the dubious distinction of having the worst record in the nation for pedestrian fatalities,” said Tri-State Transportation Campaign Policy Coordinator Nadine Lemon. [End quote]

    Now according to this Wyoming DOT statistic, 37.5% of all traffic fatalities in Wyoming in 2016 we caused by out-of-State drivers.

    Your choice of the data year 2016 wasn’t because Connecticut saw a huge 45% jump in traffic fatalities that year, was it?

    Or because “According to the National Safety Council (NSC), 2016 was the deadliest year for U.S. roads since 2007”, (even though Arizona reduced total driving fatalities from 1301 in 2006 to 962 in 2016, despite a substantial increase in population)?

    Here is some darn good data from the Insurance Institute for Highway Safety which shows that on a fatalities per 100 million miles operated standard that Alaska is the worst State in America for traffic fatalities:

    Don’t you think that perhaps among the greatest needs in NYC traffic safety would be building some pedestrian overpasses, along with mounting a pedestrian safety campaign to remind people to look both ways before they cross the street???

    One more weird question about highway safety in NYC. How much of your total fresh food supply and consumer goods supply mileage is operated in NYC and how much is operated in other States, considering that the lion’s share of your fresh produce comes from either West Coast States or Florida, and the lion’s share of your fresh and frozen meat supply comes from west of the Mississippi River too?

    Could the fact that more than 99% of your food supply mileage is operated outside of NYC contribute to higher fatality rates in other States, and tend to reduce your own fatality rate?

    How much of Great Britian’s fresh food supply mileage comes from flying so much of it in from Africa? Wouldn’t that fact tend to reduce their traffic fatality rate per-population too?

    Aren’t statistics fun and so easy to bend in-order to back-up our beliefs too?

  21. ‘Vehicle Miles’ and ‘fatalities per sq mile’ are ridiculous measurements of fatality rates. Just because our nation doubled down on unsustainable, dispersed, highway-centric sprawl, doesn’t mean we should use metrics that make it look less bad.

  22. The fact is that the vast majority of the US has been developed to a much-different standard as NYC, in-fact the two are not readily-comparable on highway needs including on speeds.

    You want to know why your fatality and injury rates are so high compared to most of the Western US, which has higher speed limits and traffic just as heavy on many roads?

    It is because Western US States generally have much wider roads and lanes than Northeastern US States, along with far fewer fixed obstructions.

    Across most of the Western US it is legal to haul an 18-foot wide mobile home as our roads here are designed so that loads this wide can be driven without straying into oncoming traffic.

    I get the idea that many big-city Eastern US drivers don’t drive as much nor have as much driving experience than most Western US drivers, for whom it might be 40 miles to the closest Walmart.

    That is why the fatality rate per square mile is a useful figure as it is a backup figure for how many miles are driven on a per-person basis. It can also show population density too if population is added to it.

    Back in 1979 the trucking industry had its worst year for total fatalities. Since then they key rate per 100 million miles

  23. Thanks for the link, there are a lot of different variables in comparing U.S. traffic fatality rates to European ones, or western states to eastern cities, or southern to northern roads, but seems obvious that we could lower our death and injury numbers considerably by applying solutions that would work here.

  24. This is a well reasoned argument, but I don’t think 18′ wide trailers can be towed on many roads in the western states. Maybe on the playa.

  25. In northern California, the police won’t enforce existing speed limits unless a recent speed survey has shown that 85% of drivers obey it, because of one judge’s ruling tossing out a speeding ticket. Since 85% of drivers tend to drive 5-10mph faster than posted speed limits, their solution is to raise the posted speed limit, then occasionally enforce that, until the race to the bottom results in most drivers going faster.

  26. Here are Colorado’s oversize load rules. Loads over 17 feet wide on most State and US primary highways require a special permit. It is interesting to note that 16-wide mobile homes do not require a special permit except on some narrow local roads.

    Most of our major roads here in Colorado have 12-14 foot lanes and paved shoulders of 8-10 feet in width plus wide clear zones without fixed obstructions. East of the mountains most of our roads are quite straight too.

    Here is a good example of a rural 2-lane here in Colorado, CO Hwy 66, with this one fairly heavily-traveled as it connects the north side of Longmont, population 93,000, with I-25 about 6 miles east of the city. With a 12-foot travel lane and a 10-foot wide paved shoulder there is 2 feet on either side of an 18-wide mobile home between the edge of the pavement and the center line. Hwy 66 has a 65 mph speed limit in this area too.,-105.021857,3a,60y,274.67h,85.87t/data=!3m6!1e1!3m4!1sKNzrpKG3Ee3qR6omY9ZA3w!2e0!7i13312!8i6656

  27. Thanks for clarifying that. That would be terrifying to be towing something that width at 65 mph. California wide loads start at around 10 feet, seem to require two CHP in front and one behind to warn oncoming traffic. I don’t know what the max is, but there aren’t many roads other than interstates with those kind of shoulders that I’ve driven on, anyways.

  28. Anthony – Ahem. Your bias is showing. If you conflagration rural and urban rates to justify your hatred of automobiles, you have lost all reasonableness. If you can’t see the difference between US and Europe you have no ability to offer any useful opinion. So keep up the statist propaganda paid for by Bloomberg.

  29. let’s try and improve people’s journeys to where they MAKE money

    then they wont mind a little congestion when they go to spend it.

  30. Roads should be designed to allow the highest possible number of vehicles at the highest possible speed, then speed limits should be set to however fast motorists want to do. Anything less is an infringement of our rights.
    James C. Walker, National Motorists Association

Leave a Reply

Your email address will not be published. Required fields are marked *


Engineering Establishment Sets Out to Purge Deviant Bikeway Designs

The National Committee on Uniform Traffic Control Devices might be the most influential group of American bike policy makers you’ve never heard of. The committee shapes street design standards in the United States to a large extent. Their recommendations become part of the Manual on Uniform Traffic Control Devices, a guide to street markings, signs, and […]