Traffic Engineers Still Rely on a Flawed 1970s Study to Reject Crosswalks

When St. Louis decided not to maintain colorful new crosswalks that residents had painted, the city’s pedestrian coordinator cited federal guidance. A 2011 FHWA memo warns that colorful designs could “create a false sense of security” for pedestrians and motorists.

Shoddy, 50-year-old research is an obstacle to grassroots street safety efforts like this fleur-de-lis crosswalk in St. Louis. Photo: Rally St. Louis

That may sound like unremarkable bureaucrat-speak, but the phrase “false sense of security” is actually a cornerstone of American engineering guidance on pedestrian safety.

You’ll find the words “false sense of security” in Washington state DOT’s crosswalk guidelines too. The city of Stockton, California, makes the same claim. The list goes on.

What gives? Well, you can trace this phrase — and the basis of some engineers’ reluctance to stripe crosswalks — to one very influential but seriously flawed study from the 1970s.

In 1972, a researcher named Bruce Herms conducted a study of crosswalk safety in San Diego. He found that intersections with marked crosswalks had higher injury rates than ones with unmarked crosswalks. He concluded that marked crosswalks should only be installed where they are “warranted” because they can give pedestrians a “false sense of security,” encouraging risky behavior.

But there were problems with the study. For one, Herms didn’t actually study why people made certain decisions at crosswalks — that “false sense of security” was just speculation on his part.

Since the Herms study, other studies have refuted his conclusions, including work produced by the FHWA. Nevertheless, the influence of his research from more than 40 years ago persists. As backward as it seems, engineers still refuse to install crosswalks on the grounds that it would harm pedestrian safety. Just a few years ago, for instance, the “false sense of security” argument was deployed to shoot down requests for midblock crossings in Los Angeles.

Bill Schultheiss, an engineer with the Toole Design Group and member of the bike and pedestrian committee of the National Committee on Uniform Traffic Control Devises, is critical of the Herms study.

“When I first came into engineering, I heard a lot about this idea of pedestrians having a false sense of security when in marked crosswalks,” he said. “And I just believed it.”

But then Schultheiss ordered a print copy of the study to review it.

“I think it was biased,” he said, like much of the older regulations. “I don’t know was it him or just the culture at the time.”

“His conclusions were terrible.”

For example, Herms found that of the pedestrians who were struck, most were hit in the middle or near the end of the crosswalk, not at the beginning. This is a pattern that suggests motorists are failing to yield to people who have already established themselves in the crosswalk, not that people are stepping off the curb inattentively.

“If you make it three quarters of the way across the street, you expect cars to stop,” Schultheiss said. “That’s the law.”

Despite the absence of evidence to back it up, the idea that crosswalks encourage pedestrians to engage in risky behavior continues to enjoy credence in the engineering profession. Official memos like FHWA’s 2011 guidance on crosswalk art repeat and endorse the idea, squashing grassroots street safety efforts.

24 thoughts on Traffic Engineers Still Rely on a Flawed 1970s Study to Reject Crosswalks

  1. Are any crashes caused by drivers running red lights or stop signs?
    Seems to me that traffic signals and stop signs give people a false sense of security. We should probably do away with them.

    Do crashes at higher speeds cause worse damage? Yes, ok, we’ll put up some speed limit signs.
    But tons of drivers will ignore the speed limit anyway. Clearly, speed limit signs cause a false sense of security.

  2. I know this is supposed to be sarcasm, but there is a grain of truth here. If stop signs or traffic signals are overused, as is the case in much of the US, particularly NYC, then they tend to lose their effect. They often ARE ignored by drivers, with the result people get killed. So while we shouldn’t do away with them completely, we should use them a lot less.

    As for speed limits, they only tend to be ignored if they’re improperly set. Legislated speed limits in general are a horrible idea, particularly on limited access highways. On urban surface streets maybe they’re OK, but they must be followed up by a street redesign which incentivizes driving at or under the legislated speed limit.

  3. Aren’t those engineers are using 1950s mentalities and 1960s traffic projection models? A 1970s study is probably really up-to-date in their world!

  4. The Herms study is simply a convenient excuse for traffic engineers to use in order to justify removing pedestrian safety infrastructure that might affect LOS. It was widely adopted and utilized because it was a cheap alternative to actually creating infrastructure that would save lives.

    Also, since traffic engineers are largely exempt from litigation, there’s no legal downside to promoting solutions that let them blame the victims for their own injuries by not “walking in a crosswalk”, even if those studies they base their conclusions on are largely discredited.

  5. Stop signs are very rarely used in The Netherlands. I would suggest that benchmarking the country with the safest streets in the world may be a good idea.

  6. Good luck getting a driver to yield in an unmarked crosswalk in NJ, much less get a cop to write a ticket when a motorist strikes someone crossing in one.

  7. Hmmm. The abstract of the FHWA study you say refutes the Herms study says,

    “… on multilane roads with traffic volumes above about 12,000 vehicles per day, having a marked crosswalk alone (without other substantial improvements) was associated with a higher pedestrian crash rate (after controlling for other site factors) compared to an unmarked crosswalk.”

    It also says crosswalks alone one two lane roads have no effect on safety.

  8. Right. Good point. We should have worded this differently and more carefully. With respect to the “false sense of security” idea, the study reviews the literature and basically concludes there isn’t enough support for that idea. That was FHWA in 2005. Then in 2011 they issue this memo on crosswalks using that exact phrase.

  9. RIP crosswalk in St. Louis.

    I hope somewhere in the future the people will see that this was short-sighted and wasn’t the right decision for it’s pedestrians.

    I’m betting the ones that made this decision use their cars to pick up that fictional gallon of milk that’s a mile away, right?

  10. We could start by looking to Massachusetts, the US state with the lowest rate of motor vehicle fatalities. You’ll find very few high-speed multi-lane surface streets and many, many uncontrolled crosswalks.

  11. Right. However, if you look at Table 1 of the 2005 study you see that on multi-lane roads (where the marked crossings are more dangerous), the pedestrian volume on the unmarked crossings is much lower, while on the two-lane roads (where marked vs. unmarked makes no difference), pedestrian volumes are about the same. This suggests that pedestrians are making choices about crossing multi-lane roads in a different way than they make choices about crossing two-lane roads, in the case of multi-lanes going out of their way to use a marked crossing. That might lead one to believe that they were gravitating to the marked crossings on busy roads out of the famous false sense of security and then crossing with less care than they would at the unmarked crossing – this is what the false sense of security hypothesis requires, that you cross with less care because it’s marked; or it could mean that the users of the unmarked crossings on multi-lane roads tended to be people who were in any case more agile and alert, or tended to use the unmarked crossings only when the road was less busy. The latter – a selection effect, not a false security effect – fits better with my personal experience.

  12. While the reasoning and outcome may be bullshit, I do think there’s something to be said for the “false sense of security” idea. About eight years ago this UConn student got killed while trying to cross the road at night; my mom is good friends with the mother of the guy who was her boyfriend at the time so I got to hear what her friends recounted about what happened. (I know the “mom’s-friend’s-son’s-girlfriend” relation seems pretty far removed, but the guy was probably going to propose to her and his parents were basically already treating her like their daughter-in-law; so it’s not quite the six degrees of Bacon that it might otherwise sound like.)

    Apparently the Carlee’s friends jaywalked across the street in the middle of the block, and she’s insisted on “crossing the street properly” and walked to the intersection and waited for the crosswalk light to change. Everything I heard about the story convinced me that she just blindly crossed the street when the light changed because she felt that the only thing that mattered was that she had the right of way.

    I think the story illustrates two things. One, I know the history about the invention of the idea of jaywalking, but this story shows you why the behavior labeled jaywalking (which is 100% legal in places like the UK) is safer. You know you have to make damn sure you can safely cross the street before you do it.

    Two, I think this story shows exactly why placing crosswalks at intersections and declaring that the only legal place to cross the street is such a dangerous idea. In the middle of a block, drivers have no real reason to be looking anywhere other than forward; at intersections, they’re trying to look in many different directions trying to make sure there’s no car traffic opposing what they’re trying to do. Which means less visibility for pedestrians.

    I’m a New Yorker and I’ve definitely done shit like cross 4-lane streets (two lanes each way) after judging “well, there’s one car coming from the left, and another crossing from the right, but if I wait for the first car to go by and walk into the middle of the street, I’ll at most have to wait just a couple of seconds for the second car to go by and then I can finish crossing.” Compare to shortly after I moved to Arlington, VA and on a three-lane street (all one way) I tried to pull the “well, there’s one car coming down the middle lane so if I step out *now* the car will safely pass by and I won’t have to break my stride” maneuver and the driver slammed his brakes so that he was stopped right in front of me, honked his horn, and was giving me this look of “are you trying to commit suicide?!”

    It’s better for everyone if there’s an expectation that drivers and pedestrians alike have to check if it’s safe to do what they want to do. I now live in Los Angeles and while I absolutely cross if the traffic moving orthogonal to my direction of movement is stopped but it’s one of those situations where the cross light never changes unless you push the beg button, I’m also a lot less of an aggressive jaywalker out here because I know the drivers are less likely to have any respect whatsoever for the fact that people might be trying to cross the street. Go figure, I make a game out of trying to see how wide a street I can successfully jaywalk across but I can still figure out when it’s a bad idea to try to do it.

  13. What a ridiculous, “engineering,” “one best way” answer to a complex problem with many acceptable solutions. Let’s list them:

    Speed tables
    Crosswalks buffered back from the curb
    Smaller turn radii
    Bricked treets that force drivers to slow down
    Bulb out curbs
    Extended reds for initial crossings
    Plastic “ped xing” bollards in the middle of the street
    Narrower streets

  14. In California the pedestrian always has the right of way. But apparently Los Angeles drivers… and police… don’t know this.

  15. Two theories here:
    (1) Because nobody’s bothered to sue the hell out of them.
    (2) “Sovereign immunity” — the way most state governments are totally exempt from litigation even if they murder babies in a ritual sacrifice to Satan on live television.

  16. Isn’t the selection effect a false security effect? Even at busy times people are willing to cross at marked crosswalks, though they’re not as safe as less busy times at unmarked crosswalks.

  17. “Herms found that of the pedestrians who were struck, most were hit in the middle or near the end of the crosswalk, not at the beginning. This is a pattern that suggests motorists are failing to yield to people who have already established themselves in the crosswalk, not that people are stepping off the curb inattentively.”

    Isn’t that exactly what the “false sense of security” is claiming? The marked crosswalk makes you think cars will yield or stop when they’re required to, but that’s just false. In an unmarked crosswalk, people think that cars won’t yield or stop (even though they’re legally required to) so they have the accurate sense of insecurity.

  18. There is a reason why you will NEVER find an “Art Crosswalk” in the countries of Germany, Denmark and the Netherlands. THEY ARE NOT, NOR DO THEY LOOK LIKE CROSSWALKS!

    I too believe the Herms study is flawed but not every “artsy fartsy” idea safe streets advocates come up with actually work. If this movement had a sense of some humility at times it would be willing to do some internal reflection and re-evaluate some of its ideas.

  19. As a majority pedestrian, I am in 100% support of marked crosswalks, because they mediate between the pedestrian and the motorist in a useful way. You can’t blindly step into a street, but it is helpful to have the markings as a guide where to step. That said, I do not like the do-it-yourself, paint-by-numbers, grass-roots crosswalks like the one in this story. Like the stop signs with which they interact, crosswalk markings should be consistent across a city. Also, I don’t like the reflective crosswalk signs in the middle of blocks unaccompanied by a stop sign.

Leave a Reply

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


Get Real — Colorful Crosswalks Aren’t Endangering Pedestrians

In the summer of 2014, residents of Tower Grove in St. Louis painted crosswalks with patterns like a fleur-de-lis to add some neighborhood character. Now city officials say the crosswalks should fade away, citing safety concerns. The order comes from new bike and pedestrian coordinator Jamie Wilson, who cites a 2011 recommendation from the Federal Highway Administration. Wilson told The Post Dispatch he […]

St. Louis to Pedestrians: Drop Dead

Here’s a great example of the wrong way to handle a tricky pedestrian crossing in your town. At the request of a local hospital, the city of St. Louis recently removed a frequently-used crosswalk for at least the next two years, apparently in conjunction with nearby construction. The city didn’t just scrub away the markings — […]

Designing City Streets to Suit 47 MPH Drivers Is a Recipe for Failure

Gravois Avenue is an important commercial street in St. Louis that also happens to be designated a state highway. It’s currently slated for a redesign, providing a huge opportunity to make the street work better for walking and biking. But unfortunately the highway-like mentality of state transportation planners persists. Alex Ihnen at NextSTL reports that Missouri DOT is using highway design […]