Yellow Journalism: Tribune Panics Over “Risky” Stoplight Timing

The Tribune is trying to brew a storm of controversy over the city's red light camera program by pointing out that Chicago, like every other city, times its yellow lights differently. Photo: Jamelah
The Tribune is trying to provoke controversy over Chicago’s red light camera program by pointing out that the city times its yellow lights differently — just like every other city. Photo: Jamelah, via Flickr

Day in and day out for at least 30 years (and perhaps for almost a century), over 3,000 stoplights all across Chicago have whirred through tens of millions of cycles the exact same way: green, then yellow for three seconds, then red. Yet today, this three second cycle was suddenly declared a public safety emergency, with the Tribune’s front page fomenting panic about the crisis posed by “risky” and “too short” yellow phases.

The Tribune, of course, has long pursued a vendetta against the automated enforcement of red lights in Chicago, consistently whining about a program that penalizes criminals who blow through stoplights with deadly consequences. In its newest episode, the newspaper assembled a cadre of experts to inveigh against the long-established three second yellow phase, and arguing for a few tenths of a second more leeway. (This isn’t the first time the Tribune has zeroed in on fractions of seconds in arguing against enforcement.) Drivers, it seems, feel as if they’ve been “ambushed” by yellow lights that work exactly the same way they’ve worked for decades.

One example the Tribune cites approvingly is Maryland, where a 2004 law lengthened the minimum yellow signal phase to 3.5 seconds. Yet the story there was all about political perception, rather than engineering standards. Frank Murphy from Baltimore’s transportation department told the Tribune, “The reason the law was passed was because it was represented that there was an ambush situation, when yellow lights were set so low – even though they had always been set at three seconds previously.”

True, some recent engineering guidance recommends that cities assume that drivers are usually speeding when approaching traffic signals, and such formulas find Chicago’s yellow signals to be on the short side. For example, Institute of Transportation Engineers’ formula recommends that for situations like a citywide standard (where actual travel speeds can’t be observed), adding 7 mph to the speed limit across the board — thus assuming that drivers citywide are traveling at 37 mph.

Moving forward with that assumption would endorse and enable speeding, which is a far cry from the Chicago Department of Transportation’s recent push to eliminate all fatalities from our streets. David Zavattero is head of traffic safety programs at CDOT, and oversees the red light camera program. He said that Chicago uses a three second yellow light because “we don’t believe it is a safe environment to be [in], basing your signal timing on a 40 mph vehicle traveling through the intersection.” Plus, Chicago’s citywide three second phase has a long history: The federal government’s Manual on Uniform Traffic Control Devices first recommended a three second minimum back in 1935, and continues to do so today.

jocelyn to traffic control
A fraction of Chicago’s traffic control boxes (like this one downtown) use computers, but most use mechanical controllers. Fewer than 16 percent are interconnected. Photo: Jocelyn Geboy, via Flickr

The main takeaway from the Tribune’s roundup of how different cities time yellow lights is that there’s no single rule for signal timing. (In fact, there are even competing definitive guidebooks.) A study from the Transportation Research Board about yellow light timing [PDF] concluded that nearly every city calculates signal times differently, and even the ITE’s formula ultimately leaves final decisions to local authorities. That formula recommends a 3.2 second yellow phase given Chicago’s citywide 30 mph speed limit — but that 0.2 second difference is literally less than a blink of an eye.

Several studies have indeed concluded that increasing the yellow light duration can reduce crashes, although neither ITE and MUTCD have yet issued official guidance to that end. It’s encouraging, then, that CDOT commissioner Rebekah Scheinfeld is open to using different standards and assumptions for timing traffic signals, and for operating red light cameras.

In a perfect world, CDOT could meet every transportation engineering standard, and the city could replace its antiquated mechanical traffic lights — almost 85 percent of which can’t be reprogrammed remotely. With just a few mouse clicks, new computerized signals could adjust yellow light phases for the speed limits of various streets, create widespread signal priority for buses, give bicyclists priority on “spoke routes” with “green waves,” and use signals to keep people from speeding on dangerous streets. In that perfect world, all crosswalks would be brightly zebra-striped, all corridors would have bike facilities, all ‘L’ stations would be wheelchair-accessible, and all bridges would be structurally sound.

We, of course, don’t live in that perfect world. Of all the battles to pick to get there, it’s odd that the Tribune should fixate on the lowly yellow lights that we all take for granted.

Updated on December 24, 2014 with additional information from CDOT and clarifications regarding MUTCD compliance,  the ITE formula, signal interconnects, and historical background.

Note to readers: Streetsblog will resume publication on Friday. Have a great holiday.

  • Chris Chaten

    A car traveling 30mph is going 44 feet per second or ~9 feet in the 0.2 seconds. That’s hardly insignificant. I’m not sure what the fight is about here. Yellows following guidelines improve overall safety.

  • CL

    “Several studies have indeed concluded that increasing the yellow light duration can reduce crashes.”

    I don’t understand why this isn’t the most important thing when it comes to the issue of yellow lights. Longer yellow lights = fewer crashes. Even you guys admit it. But you’d rather make a point about speeding than give more people time to react even though you KNOW the result of your position is more accidents. It just seems very hypocritical given this blog’s goal of safety.

    Maybe it shouldn’t be the Tribune’s only battle — personally, I would like them to take up the cause of crosswalk enforcement — but this is a valid issue, especially given the finding that there has been an increase in rear-end collisions at camera intersections. Crashes are bad. You can target speeding in other ways (road engineering) while also increasing safety in every possible way, which means more time to react to a changing light. This increases safety for drivers going the speed limit, not just for speeders.

    To me, the difference when I’m driving in Evanston is noticeable. It’s usually rush hour, so I’d be lucky to even be going the speed limit let alone speeding — but I still feel a lot more comfortable with the longer yellow. I don’t have to make any split-second decisions or stop suddenly. I’m calmer at intersections, and I’m not constantly trying to watch the pedestrian countdown clocks to find out if I should be slowing down.

  • CDOT is saying they are following guidelines. None of the experts the Tribune interviewed have said they aren’t. Every city the Tribune pointed out is following the same guidelines, but they’ve all come up with different results on what the yellow light timing should be.

    Is there a one, true duration? Nope.

  • Chitown88

    Can you please explain YOUR vendetta against the Chicago tribune for reporting on what is a corrupt, controversial program?
    Why do you constantly berate them, when you do the exact same thing? It’s like you want more accidents and you want the streets to be unsafe. You would actually advocate for shorter yellow lights? You call people who go through and intersection 0.1 of a second a criminal?
    You support the city sending out tickets after they used illegally short yellows?
    Please, explain yourself.

    Your vendetta just doesn’t make sense.

  • It’s not as simple as saying “the yellow lights should be longer”. The Tribune story’s wasn’t even about saying the yellow lights should be longer because longer yellow lights lead to fewer instances of red light-running and fewer crashes and injuries.

    Longer all-red phases also lead to fewer crashes and injuries.

    Their writers, their experts, and I, are not in a position to tell CDOT what the yellow light should be…at 1 intersection, at the 190 with red light cameras, or at 3,000 intersections.

  • Chris Chaten

    Basically everyone quoted in the Trib article and the WBEZ article you linked to think the Chicago method is too short, and erring on the long side is more effective at reducing crashes.

    “Robert Seyfried, a professor of engineering at Northwestern University’s Center for Public Safety, edited the most recent edition of the ITE’s Traffic Control Devices handbook.

    ‘That is clearly not following the ITE recommended practices,’ he said of Chicago’s practice of changing the deceleration value.”

  • Hugh

    CDOT knows what’s best, Steve???

  • Hugh

    Yeah, what he said, Steve. What’s the vendetta?

  • Marc Dreyfuss

    The Tribune’s push to remove cameras is ridiculous. If the lights are timed poorly though, that’s a separate issue that should be dealt with. Yellow lights in Chicago ARE too short, which partly explains the cultural normative question of “How many cars are allowed to go after the light turns red?”

    This should be studied, but as a transportation planner (not engineer), I thoroughly reject your suggestion that safety features of any kind should be set up in a way that assumes people follow the law to a ‘t’, rather than set up in a way that facilitates safety under real conditions. Most people drive above the speed limit. If you want them to go slower: narrow the streets, add stops signs, lower the speed limit. In Maryland they also have speed cameras, but they only get you if you’re going at least 12 mph over the limit. Short yellow lights just makes things dangerous.

    So what to do? 3.5-4 sec yellow. Delayed “all red” to clear the intersection, pedestrian priority by giving them the walk sign 1-2 seconds before the cars, and more NO TURN ON RED signs. Regulating traffic will reduce accidents, not hyperbolic nonsense.

  • Guest

    And we all know that ITE is the paragon of accuracy and rigorous scientific estimates…

    http://usa.streetsblog.org/2014/12/12/whoops-how-planners-and-engineers-badly-overestimate-car-traffic/

  • While continuing to research I found this statement from the TRB study that we linked to, which shows that the 10 and 12 feet per second per second deceleration rates that ITE recommends and CDOT uses, respectively, are lower than what the rate drivers were observed decelerating.

    “The findings of a 2007 experimental study supported a significant relationship between deceleration rate and time to stop line to age. Deceleration rate decreased as drivers were farther from the stop line. The mean deceleration rate for 18- to 35-year-old drivers was 14.4 ft/s2, compared to 12.5 ft/s2 for 55- to 64-year-old drivers and 12.3 ft/s2 for 65-year-old and older drivers (ref. 22).”

    CDOT uses 11.2 feet per second per second, a deceleration rate identified by the AASHTO for sight line distance stopping. This is because, according to CDOT, “In Chicago, traffic engineers must consider high roadway volumes and intersections that are spaced as closely as 300 feet together. These factors cause drivers to decelerate at a higher rate.”

    If CDOT used the 12.3 feet per second per second figure (representing senior citizens, who according to that study need the most distance to stop) in the ITE formula the yellow light duration would be 2.8 seconds.

  • Chris Chaten

    If you use a baseline that isn’t recommended by anyone you can get whatever duration you want. The standard for yellow lights isn’t the mean. Most recommend using an 85% threshold. That study also doesn’t accommodate CTA buses or 30 MPH speed limits.

    I’m not sure what your goal is. At 2.8 seconds, a bike traveling 10 MPH won’t clear most intersections. That doesn’t promote a low stress biking environment.

  • Mark

    RL cameras should only be used in conjunction with countdown timers.

  • Brendan Kevenides

    Three second yellows are a problem for Chicago bicyclists. At many intersections around the city a cyclist may enter on a green and see the light cycle all the way to red before making it across. The result can be tragic, especially if a motorist is timing the light. Of course, adding fractions of a second to yellow lights will not solve that problem. We would like to see yellow light extended by seconds make our streets safer for cyclists. http://www.mybikeadvocate.com/2014/06/changing-lights-dangerous-time.html

  • Anyone that is genuinely concerned about accident rates and public safety would support investments in transit and alternative transportation, not investments that make it easier or more convenient for more cars to go faster in urban areas.

    There’s an inherent disconnect in using accident rates and public safety as a basis for arguing about red light cameras, yellow light timing, etc.

    Fewer cars traveling at slower speeds are safer than more cars traveling at higher speeds. It seems speed and red light cameras encourage compliance with traffic laws, though administration of such programs does need to be fair.

  • Fred

    I don’t think extra long yellows are a good solution to this problem. As long as yellow means “gun it!”, adding 3 seconds of yellow adds 3 seconds of excessively speeding vehicles flying past cyclists and on the other side of the intersection. All-red periods are a far better solution to this problem. A 3.1 second yellow followed by 2 seconds of all red seems FAR safer to me than 5 seconds of yellow.

  • Anne A

    One of the worst intersections in the city for crashes and poor sight lines is Pershing and Western. I passed through that intersection twice yesterday. At least for north- and southbound traffic, the yellow phase is significantly less than 3 seconds, which certainly doesn’t help safety at this location. Due to the terrible sight lines, I’d love to see a yellow phase of 4 or 5 seconds at this intersection.

  • Hugh

    May i ask, what is your evidence for the “Tribune’s push to remove cameras…”? Would you perhaps agree that the Tribune’s editorial page is the best indicator of the Tribune’s motives? From 12/23/14: “…if the program is truly about safety, then the cameras should go where they will prevent the most accidents.”

  • Hugh

    One would hope that UIC CUPPA while training young minds to criticize studies also stresses the fundamental principal of criticizing a study without criticizing the authors.

  • Bernd

    ITE and MUTCD standards were changed decades ago to allow for shortened yellow signal times and then changed the recommendation to enforcement. It was this change that created the red light running problem and the enforcement that followed. This was covered in Dick Armey’s congressional report on red light cameras. Yes. That long ago. Every place where by court order or legislation where red light camera intersections were forced to be fixed the cameras became unprofitable.

    The Tribune’s studies on red light cameras are consistent with all the fairly done studies every where else. Only those that were purposely set up in favor of red light cameras by those in the revenue stream and those who are just plain against driving have shown anything different. For a decade people have dismissed the overwhelming evidence because it wasn’t from Chicago. Now we have learned that Chicago is not any place special, which those of us without an ulterior motive have known all along.

    The simple fact is that red light cameras make their money in the first half second. This means we are dealing with people making simple judgment errors. As the yellow signal increases in duration these errors drops off in the same proportion. The dilemma zone gets smaller as the yellow time increases until it no longer exists. This is why a longer yellow, even a tiny bit longer, decimates the profitability of the camera system.

    At this point only those with motives other than safety can possibly support short yellow times and red light cameras. If safety is the goal then what difference does the yellow signal length matter? Just lengthen it and keep the camera. If the theory that longer yellows does nothing were true then there is absolutely no reason to fight this. No reason to write articles like the one above. Just do things right. Have the camera if you want it, but do things right. The very fact that doing things right is opposed in favor of red light cameras shows that other motives are in play and they know the cameras will become uneconomical.

  • Wewilliewinkleman

    Another comparison should be made where intersections that have high volume of traffic or high incidents of crashes with no red light cams but with pedestrian count down clocks. If the long term effect of installing a count down clock has the same, or a greater effect of reducing traffic accidents versus the period befor the count down clock was installed red light cameras may not be be necessary unless the only reason is to collect fines.

  • Bernd

    That idea was rejected by the city council years ago. ( http://theexpiredmeter.com/2009/02/alderman-introduces-countdown-timers-at-red-light-camera-intersections/ ) Watching the ped signal, even before countdown timers, before red light cameras, was something many motorists learned to do in Chicago because of the short yellow signals. It’s a sign that things have wrong for decades.

  • Bernd

    The problem is that one must pay for the ITE standard so reporters do not purchase it to read it. There are various factors that go into what timing should be. Speed, road width, and more. However, the basic way of setting a yellow signal changed over time which requires looking at older ITE standards to understand how the problem was created in the first place. Thankfully there is an easy to find congressional report that summarizes it:

    http://www.motorists.org/red-light-cameras/armeyreport.pdf

    In 1976 the standards were for much longer yellow signals. This was changed in the 1980s. In 1994 the words ‘use enforcement’ were added. The detailed discussion of these changes starts on page 12 of the pdf linked above.

    Not only is the current standard often cheated upon across the USA for red light cameras, the standard itself is inadequate. All we need to do is go back to the 1976 standard if we want safety.

  • Fred Rubble

    At each instance making driving more expensive and/or difficult will take priority over the needs of safety, bicycling, or transit. This should indicate to the savvy reader what streetsblog’s agenda is.

  • Wewilliewinkleman

    And it also encourages bikers not to follow predictable traffic patterns which makes driving more stressful, and encourages resentment of bikers.

  • giovanni

    The oddest thing about the whole debate is that both sides have used an absurd statistical premise for measuring the effect of the cameras – changing crash rates at camera corners vs. other corners in Chicago. The problem with this premise is that most people do not have a omnipresent mental map of the red light cameras in the city, stopping suddenly and erratically at those corners with cameras and whizzing through other corners. Instead, they occasionally remember a camera at a given intersection, but the existence of a few cameras also changes behavior citywide, as drivers realize they are more likely to be caught blowing a red light, and tend to make a slightly different decision when a light turns yellow.

    When the Trib adjusted the dramatic decline in rear-end crashes at camera intersections downward because their “control group” of non-camera intersections also declined, they turned an argument FOR the cameras into a pseudo-statistical argument against them.

  • paulcycles

    I’d love to get an extra second at notorious blind, busy, or really wide intersections like Anne mentions at Ashland & Pershing/39th St. I’ve had near misses since 1968 on my bikes, cars, delivery trucks, & esp. When I drove for the CTA back in the day. Pershin RD has been a how fast can I drink & drive strip since before my birth in 1952!
    It was much safer in the days of the Overpass on Ashland circa ~1967 to ~2000, BUT
    bikes & commercial trucks and busses NOT so much…

  • Hugh

    “…both sides…” If there are 2 sides in this debate it seems to me they are not the Tribune vs. right-thinking persons but rather an administration that believes it is above accountability playing revenue games with public safety.

  • Hugh

    Yes, the Tribune reported that accident are down overall. This is more than the City acknowledged in their defence of the red-light program at in public hearing in City Council chambers. And yes, AMONG the factors reported by the Tribune as contributing to that overall decline was the “halo effect,” others being the economic recession and improved automotive safety. But I don’t find the Tribune argument against the cameras you mention.

    The City touts the halo effect as a FEATURE of their program. For example, intersections with cameras on just 2 of 4 approaches have signs on all 4 approaches.

    The Tribune study is not perfect. The Emanuel administration’s defence of their red light program to date has been deliberately misleading and embarrassingly incompetent, well beyond the point of compromising public safety.

  • Shlabotnik

    thanks for your insightful comments on this topic, here and elsewhere.

  • giovanni

    Right. You didn’t see it. Because they didn’t acknowledge it in their article. You have to read the study to realize the ridiculous methodology. It’s only by fudging the numbers that you can pretend that t-bone crashes haven’t fallen substantially, far more than the increase in minor rear-end crashes.

    The halo effect isn’t about 2 of 4 approaches. It’s about the whole city, because as I mentioned, most people don’t keep track of where cameras are, they just know that they’re out there and drive somewhat more cautiously everywhere as a result

  • Tim or Timmer or Timothy

    I only drive about a half dozen times a year when I rent a Zipcar. Last year, I received a notice of a violation supposedly occurring at an intersection with a red light camera. With the violation notice received several weeks after the fact, all memories of driving that day fell into generic memories, and I certainly had no recollection of the intersection in question. (And the camera footage they had on the web site wasn’t helpful either.) If the goal of the ticket is to modify my driving behavior, it doesn’t seem helpful to send the ticket after enough time has passed that I don’t recall my driving behavior any longer.

  • Fred Rubble

    Automated enforcement is designed to profit from normal behavior by taking advantage of errors or deliberate mis-engineering in the road system along with ‘gotchas’ like not coming to a complete stop or not doing a textbook double stop for a right on red. There is no ‘better’ driving to be encouraged by RLC tickets, only driving to avoid RLC tickets by for instance never turning right on red at RLC intersections.

  • MF

    Streetsblog accusing the Trib of yellow journalism looks like a case of the pot calling the kettle black — or, i suppose, yellow, in this case

  • ChicagoCyclist

    I am 100% for red light cameras and for speed cameras too. These
    technologies represent fair, objective, automated, efficient
    enforcement. The safety benefits are well-established. The only real
    arguments against these cameras are basically: “I want
    to be able to run red lights without receiving a ticket and I want others
    to be able to too,” and “I want to be able to speed without receiving a
    ticket and I want others to be able to too.” That said, the fact that “there’s currently no
    national agreement on what the standard [for the duration of yellow
    lights]
    should be, and there’s no indication that Chicago’s current yellow
    light timing is unsafe” does NOT represent a sound argument against
    increasing the
    yellow light duration from 3 to 3.5 seconds. Similar to the situation with climate
    change, the LARGE MAJORITY of engineers around the country and the large
    majority of jurisdictions around the county have yellow light durations
    longer than 3 seconds. Why not follow suit here in Chicago? The only
    real
    argument against doing so, which I can think of, is: “We don’t want to
    slow automobile throughput.” That very old, out-of-date
    automobile-centric argument is
    NOT valid in a multimodal, urban environment like Chicago. Nor is it
    consistent or in keeping with the Complete Streets policy / approach
    that the City and CDOT have adopted, which (generally and in most
    locations/situations) prioritizes pedestrians and bicyclists (and
    transit service/users, which includes of course pedestrians) OVER automobile travel.
    Why? Because they (peds and bikes) are the most vulnerable roadway users. The
    idea is the same as the spirit behind Vision Zero, which the City has also
    officially adopted — namely, the goal of reducing fatalities. In my opinion, Chicago should have red light and speed cameras almost everywhere, and should ALSO increase yellow light duration to 3.25 or 3.5 seconds! Why not?

ALSO ON STREETSBLOG

CDOT Tells Council Improved Red Light Cameras Might Mean More Tickets

|
Chicago Department of Transportation commissioner Rebekah Scheinfeld told the City Council’s Committee on Pedestrian & Traffic Safety that equipment problems that have plagued the city’s red light camera enforcement program have mostly been resolved. Although some aldermen were there to pounce upon the media ruckus around the cameras, other aldermen understood that red light cameras can […]