Stop Sign Compliance

In conjunction with the “Bikes Safe at Stop Sign” bill recently introduced in the Arizona legislature, I began to wonder more generally about stop signs in general (all traffic, not just bicycles).


As usual, the first stop (pun?) is to look at ARS Title 28 to see what’s what with stopping.

The stop & yield sign rule is §28-855. The rule says everyone has to STOP, and goes on to specify exactly where; depending on whether or not there is a crosswalk, stop line, etc. There is no wiggle room.  (for what to do afterwards, see §28-773 — thanks eric)

I also learned that a driver approaching a driveway “within a business or residence district” (i.e. not rural) must, in effect, treat it just like a stop sign (regardless of whatever signage may or may not be there). The driveway law is  §28-856. And, satisfyingly, it does specify what a driver must do after stopping.

In contrast, the yield sign rule seems much more, well, reasonable: “slow down… to a speed reasonable… and shall yield the right-of-way”. There is no arbitrary standard.

As a confusing side story:  §28-1501 has, among others, the definition ” ‘Stop’, if required, means complete cessation from movement” ; but strangely the section specifically applies “to this chapter” (referring to chapter 5 “Penalties and Procedures for Vehicle Violations”. Rules of the road like 855 and 856 are in Chapter 3.


I was interested in finding out what proportion of drivers actually stop when there is no conflicting traffic. Traffic engineers apparently categorize behavior into 4 types. The first two comprise legal stops; Voluntary Full Stop (VFS) which is stopping when there is no conflicting traffic, as distinguished from Stopped by Traffic. The other two would be illegal; Almost Stopped, and Non-Stopping.

There was much in the Behavioral / Social Sciences literature on this.

Feest (1968), Lebbon (2007),  McKevlie (1986), Trinkaus (1997) observed a VFS rate of only 15%, 4.6%, 24.1%, and 1% respectively. VFS (Voluntary Full Stop) means fully stopped other than being stopped in order to yield to any other traffic.

So it appears that somewhere between the majority and the vast majority of drivers do not stop when there is no conflicting traffic. This comports with my anecdotal observations; see my page of anecdotal videos showing very high non-compliance rates.

From traffic engineering literature, the general conclusion is that compliance isn’t the problem, driver-error is.  E.g. “In short, the results of many previous studies suggest that accidents at twoway stop controlled intersections are more closely related to driver error, such as failure to accurately judge the speed of major roadway vehicles, than to roadway geometry, sight distance and driver compliance with traffic control devices” (Stokes 2000)


Box and Oppenlander, 1976, (a book, i’m guessing a Traffic engineering textbook)  This worksheet defines 5 levels of compliance, note that only the first two are legal, the “practically” is an expedient, 0 mph should be categorized as either VFS or Stopped by traffic, but hard to discern in the field:

  • Voluntary Full Stop
  • Stopped by Traffic
  • Practically Stopped – 0 to 3mph
  • Non-Stopping

Bretherton, MW 1999.  Multi-way Stops – The Research Shows the MUTCD is Correct! . ( retrieved full text Nov 28, 2015). This paper focused on why it is undesirable to use stop signs for the purpose of speed-control, but it also has loads of references.

Feest, Johannes (1968) Compliance with Legal Regulations: Observation of Stop Sign Behavior. Law & Society Review, Vol. 2, No. 3 (May, 1968), pp. 447-461.This is a fun paper. Feest holds up the law about requirement of fully stopping at a stop sign when there is no cross traffic as an example of an “Unstigmatic regulation”, i.e. a rule that is frequently flaunted with little consequence. Feest found in the absence of cross-traffic “the number of people who strictly comply with the formal legal regulation is about 15 per cent”, and that the large majority made either a “rolling stop” (58%, 2-10mph) or “half-stop” (22%, 10-20mph). These categories taken from a pamphlet used for training law enforcement officers.

Lebbon, A.R., et al (2007) Evaluating the Effects of Traffic on Driver Stopping and Turn Signal Use at a Stop Sign: A Systematic Replication. Journal of Organizational Behavior Management, Vol. 27(2) 2007.

McKelvie, S. J. (1986). An opinion survey and longitudinal study of driver behavior at stop signs. Canadian Journal of Behavioral Science, Vol. 18(1), pp. 75-85. “most drivers stopped when another vehicle was in the vicinity and most only slowed when the road seemed clear. It is concluded that stop sign behaviour reflects risky decision-making rather than adherence to the law…”. For observational result of 24.1% VFS, see Table 3. One hiccup is they defined what they termed a Complete Stop as stationary for 2 seconds, with is supra legal.

McKelvie, S. J., & Schamer, L.A. (1988). Effects of night, passengers, and sex on driver behavior at stop signs. Journal of Social Psychology, 128(5)(Oct 1988), pp. 685-690. “Despite the legal requirement that drivers come to a complete halt at stop signs… absence of other vehicles lowers (compliance) to 15%”

Mounce, JM. (1981) . Driver Compliance with stop-sign control at low-volume intersections Transportation Research Record(TRB) No. 808. pp 30-37. To paraphrase from the abstract: compliance (a full stop, whether you need to or not) is high when there is a lot of conflicting traffic, and low when there is little traffic. And in addition, if the sight line is good, compliance is lower. (which of course makes perfect sense) “The results from 2830 observations at 66 intersections indicated that the violation rate decreases with increasing major-roadway volume and is significantly high (p < 0.001) up to the average-daily-traffic (ADT) level of 2000 and significantly low (p < 0.001) above the ADT level of 5000-6000. An interaction effect between major-roadway volume and minor-roadway sight distance results in a violation rate that is significantly higher (P < 0.05) when sight is unrestricted than it is when sight is restricted”

Portland Bureau of Transportation; in numbers collected in 2006 and 2007 quoted at an Oregonian article “While the study did indeed show that bicyclists come to a complete stop only 7 percent of the time, it also showed that motorists stop completely only 22 percent of the time.”.

Stokes, RW et al, (2000)  Analysis of Rural intersection Accidents Caused by Stop Sign Violation. Report No. K-TRAN: KSU-98-6. Had this to say about stop sign compliance (emphasis added): “The results of this study (and previous studies) suggest that disregard for Stop signs and other traffic control devices is not the primary cause of accidents at rural two-way stop controlled intersections. The majority of the accidents appear to be due to drivers who enter the major roadway and do not (or cannot) accelerate quickly enough to avoid being struck by major roadway vehicles. This would suggest that drivers on the minor roadway either did not see oncoming vehicles or failed to accurately estimate the speeds of oncoming vehicles on the major roadway” … “In short, the results of many previous studies suggest that accidents at twoway stop controlled intersections are more closely related to driver error, such as failure to accurately judge the speed of major roadway vehicles, than to roadway geometry, sight distance and driver compliance with traffic control devices

Trinkaus, J 1997 “Stop Sign Compliance: A Final Look“, Perceptual and Motor Skills, 1997,235.217-218 This was done with 4 specific, residential intersections over a long time period. All of the stops were considered to have been added to control speeding, and discourage cut-through traffic; i.e. as opposed to a traffic engineering reasons (there was no “warrant” for their use), and all were at T-intersections. It’s a longstanding and well understood phenomenon that drivers tend not to comply with unwarranted traffic controls.  “…it appears that motorists’ compliance
with stop signs in settings such as observed here for all intents
and purposes is now nonexistent… During a 17-yr. period full stops declined from about 37% to 1%“.

Here is what the MUTCD says about stop signs. Only tangentially related to the issue at hand, there is a lot of controversy about placing so-called unnecessary stops (e.g. placing stops on the busier road, or adding extra all-way stops) — MUTCD guidance states flatly “STOP signs should not be used for speed control.”

In a paper that surveys various traffic calming measures [Traffic Calming: State of the Practice Reid Ewing, August 1999], makes reference to data from Department of Transportation, “Brentford Lane—Stop Sign Compliance Study,” Gwinnett County, GA, September 1997. A pie chart is presented for compliance with all-way stops: 59.3% Rolling Stop, 22.2% Slowed Down, 11.1 Complete Disregard, and a grand total of 7.4% COMPLETE STOP.

Newer Roundups

I haven’t looked into this for awhile, this much newer reference Scofflaw bicycling: Illegal but rational  [.pdf] appears in a 2018 Journal of Transport and Land Used. It has a newer roundup of various illegal behaviors for everyone (bicyclists, peds, and drivers). The “but rational” part is more like rationalization, but they do point out the heart of the problem, that is drivers tend to disregard their own law-breaking, while admonishing bicyclists:

…Taking the lane refers to the vehicular cycling practice of traveling near the center of a travel lane in order to control the lane, avoid the door zone of parked cars, and prevent cars from passing too closely. While legal, taking the lane is often interpreted by drivers as rude or reckless.

…When including driving and pedestrian scenario responses—such as how often respondents drive over the speed limit or jaywalk—100% of our sample population admitted to some form of law-breaking in the transportation system (i.e., everybody is technically a criminal). When disaggregating by mode, 95.87% of bicyclists, 97.90% of pedestrians, and nearly all drivers (99.97%) selected responses that would be considered illegal …

When it comes to rule-breaking bicyclists, one popular opinion is that if bicyclists want to be taken seriously as road users, they need to obey the rules of the road like everyone else. Our survey results and the literature review both suggest that drivers break the rules of the road just as much, if not more, than bicyclists.



13 thoughts on “Stop Sign Compliance”

  1. I am in favor of the stop sign being treated as a yield sign for cyclists. Cyclists have a heightened awareness and unobstructed view (no blind spots) of interesections and tend to know when there are vehicles present or not.

    Cyclists have the added burden of unclipping shoes to put a foot down, then sometimes fumbling during re-clip when starting up in the intersection. By not unclipping, a cyclist can put their attention on the intersection where it belongs and not on their equipment.

    Cyclists may tend to use more major artories rather than neighborhood roads because of the excessive use of stopsigns in the neighborhoods. This effects the flow of traffic.

    Cyclists have to unclip several times in a round about that has stop signs, thus making it overly burdensom to negotiate. The City of Tucson is implementing a poliy of establishing Yield signs (see Tom Thivener, Tucson DOT BIke/Ped Coordinator) so as to relieve this problem. However retrofit is very costly. This new law alleviates the cost of retrofit.

    I disagree with the above blog that there is no guidance of what to do after a stop. ARS 28-855 requires a motorist to remain stopped and not proceed until safe to do so. Thus the motorist who stops must also yield. That is all we are asking of cylists.

    An automobile driver does not have to open a door and put a foot down on the roadway when stopping so it is less burdensom to auto drivers.

    A 150 hp engine produces about 100,000 watts of power. The average cyclist produces about 100 watts of power. Thus starting up is very taxing to a cyclist and not so to a motorist.

    Cyclists run the risk of falling over if they cannot unclip fast enough at a stop sign. 50 % of bicyclist crashes are single rider crashes at 0 mph (falling over). Helmets are made for the fall over crash and not made for other types of crashes simply because that is the most common. This law will reduce rider crashes from falling over and may lead to better helmet technology if the demand for a different kind of protection increases (i.e., helmets that try to work for higher speed collisions if that is the majority of crash data left after this new law goes into effect.)

    The new law does not promote wild and improper treatment of stop signs because of the prima facia evidence burden shifting language. A yield is still required.

    Do pedestrians have to stop at stop signs? Maybe. They don’t. They only stop when traffic is present. Shouldn’t slow moving cyclists be treated similarly to slow moving pedestrians in this situation?

    Again, I am in favor of Dan Patterson’s proposal.

  2. I stole this comment from a comment over at TBL

    # bb Says:
    January 8th, 2010 at 12:18 am

    Very passionate about this subject.

    Here is a T intersection same thing going on

    Here is a No right turn on red

    Here is my article on the subject

    “Yet it baffles me why we can’t change this law to reflect laws already in place. It is okay for a cyclist to use their own judgment when or when not to ride to the right. Yet when coming to a stop sign or stop light suddenly they can’t use their own judgment?”

  3. some pedestrian fatalities in late 2009 have caused a ruckus in Philadelphia, covered in Bob Mionske’s blog: Philly Pushback. In any event, someone there produced a youtube video from Philadelphia showing the almost complete lack of compliance at a stop sign

  4. The book reference sounds really interesting; but i can’t find it googling around… i do see this paper, The Psychology of Drivers.
    Its full text is there, and the author, a Herbert Stack, quotes from Brody’s book and appears to have been a collaborator with him on other work. Tom Vanderbilt pulled out this quote:
    “problem drivers are problem people; or rather, people with problems, including problems of which they often are not aware”

    Here’s an oldie but maybe a goodie? Driver Behavior and Accidents Jean K. Boek

  5. Thread from discussion group:
    Re: national project to support appropriate recognition of yield control
    Dwight Kingsbury 07/21/2008
    It’s no secret that many low-volume, nominally “stop-controlled” intersections in the US operate effectively (more or less) as yield-controlled intersections. Drivers on the stop-controlled approaches reduce speed to about 5-10 mph, scan for conflicting traffic, but tend to stop only when potentially conflicting traffic is visible….

  6. List of statues that require a STRICT stop (complete cessation of motion), no if’s and’s or but’s:

    Right on red 28-645(3)(b)
    Plain-old red signal 28-645(3)(a)
    Stop signs 28-855(B)
    driveway egress 28-856(1)
    flashing red signal 28-647(1)

    The plain-old red signal law is widely obeyed, the other are widely disobeyed (conditions permitting, e.g. sight lines and conflicting traffic)

    sort of but not really:
    school bus stop, drivers merely need to not pass the stop sign…. if you slow down it’s okay as you don’t go past.
    Same idea for school crossings, law says stop, but what that seems to mean is you must stop before entering the xwalk; it’s okay to simply slow down such that you never get to the xwalk when a kid’s in it.
    some goofy stuff about RR tracks but that only applies to special vehicle types (school buses, hazardous cargo)

  7. This morning 11/25/2015, while riding my bicycle I was sited by Pima County Sheriff deputy officer Judd for rolling through a stop sign at the four-way stop at Whitehouse Rd and Continental Rd in Green Valley. There was no cross traffic nor turning traffic. It appears I will have to go to a learning class (which at one time I taught) or pay a $187 fine. I notice that many states let bicyclists treat stop signs as yield signs. I am approaching 60,000 bicycle miles without a vehicle/bicycle involvement. I think that for Tucson/Pima County to remain a bike friendly place there needs to be some PR training of their patrol officers to encourage cycling. To be fair, why not cite vehicles that wander into marked bike lanes?

  8. The TPCBAC (Tucson-Pima County Bicycle Advisory Committee) has repeatedly asked that law enforcement focus on wrong way or opposite to adjacent traffic riders and driver pull outs (not looking to the right) because this accounts for 28 percent of all bike/auto crashes in Tucson. We have asked that they not be so concerned about stop signs because this accounts for 2.6 percent of bike/auto crashes.

    Law Enforcement has refused to follow our requests. TPD’s rationale is that riders who violate the law are going to do so in many ways so why not get them at a funnel point like a stop sign. Not sure I buy that excuse, but it is what we heard.

Comments are closed.