Tuesday, June 28, 2011

Are Right Turns on Red Dangerous for Pedestrians?

As part of the ongoing red light camera debate, we've been hearing assertions that right turn on red (RTOR) violations aren't that dangerous, because collisions due to RTOR violations are generally less severe than other types of crashes. But does this hold true for crashes involving pedestrians? And for that matter, is it true at all? (I'm always skeptical of broad statements--including my own--made without proper references.)

I took a stroll around the internet in hopes of answering those questions, and here's what I found:

The push to allow RTORs began in the mid-1970s as part of a national effort, sparked by the 1975 Energy Policy and Conservation Act, to conserve energy. By the end of the decade, most states had adopted laws allowing the RTOR in most locations. One of the earlier studies examining the safety effects of the new laws was published in late 1980. Adoption of right turn on red: Effects on crashes at signalized intersections showed an increase in crashes when RTORs were allowed, with a 60 percent increase in crashes involving pedestrians (though this large percentage increase could be due, in part, to the relatively low number of right-turn crashes involving pedestrians).

A slightly more recent study (1994) from the National Highway Traffic Safety Administration showed that while RTOR crashes represent a very small number of collisions at signalized intersections overall (0.4 percent), these crashes frequently involve bicyclists and pedestrians (22 percent of all RTOR crashes). And although the analysis shows that RTOR crashes rarely result in fatalities (less than one percent of all fatal ped/bike crashes involved RTOR), when a cyclist or pedestrian is involved in a RTOR crash they are nearly always injured.

A1996 evaluation of Pedestrian and Bicycle Crash Types of the Early 1990s from the Federal Highway Administration had similar results, showing that only about two percent of pedestrian crashes involved right turns on red.

More recently, in 2002, some folks in San Francisco conducted a Safety Evaluation of Right Turn on Red within their city. As with the earlier studies, this one found that pedestrian collisions involving RTOR represented only about one percent of all pedestrian crashes in the city.

Finally, you may have noticed that Jay Beeber of Safer Streets LA pointed out in his comments on an earlier post that SWITRS data from 2009 that shows only seven pedestrian collisions out of about 2,700 in LA were due to rolling right turns (I count four, but this could be due to my lack of familiarity with SWITRS data codes).

So what does all this mean? Are RTORs really a problem when they represent so few collisions--and if not, should we stop enforcing RTORs with red light cameras?

What I find most compelling out of all this research is the finding from NHTSA that RTOR crashes involving pedestrians and bicyclists nearly always result in injury (of course, the same could probably be said of most pedestrian crashes). To me, this means RTORs are one--of many--safety problems worth addressing. On the other hand, the low overall number of RTOR crashes suggests that in many locations allowing RTORs is not a problem.

If I ruled the transportation world (oh, if only), my suggestion would be to identify those locations where the RTOR has the biggest potential to impact pedestrian safety--namely, intersections with high amounts of pedestrian traffic. At these locations the RTOR move should be prohibited entirely, and/or enforcement should be increased (via red light cameras, or other means).  This would allow us to focus our limited resources on the intersections where they matter most--and hopefully make a noticeable improvement in our city's pedestrian safety.


  1. I'm one of those people who think RTOR should just be banned, anywhere and everywhere. There may be some circumstances where I'd be OK with police officers looking the other way despite a violation, but it should never be legal. I've had too much experience as a pedestrian avoiding collisions with drivers who aren't looking in the direction they're driving -- they drive directly into my path while looking left. The fact that I haven't been hit is entirely because I expect most drivers turning right on red to behave irresponsibly and adjust my behavior accordingly (and resentfully). So I don't think #s of collisions or injuries tell us anything about how safe the practice is. I also don't think this is an area where any risk is worthwhile -- no benefit to drivers is worth any risk to pedestrians in my mind.

    1. I agree about the dangers too. I agree that human lives are too valuable. People walk at different speeds. If it's safer, it can save a lot of grief. If they ban RTOR, driving could be easier because there's less chance of drivers behind honking when they think there's a safe gap. I was narrowly missed when I stepped out of the curb with the walk signal. I had to step back to prevent being hit. One was a fast right turn and the other a fast right turn. In both cases, the drivers had a green light. Pedestrian safety is more than just look both ways before crossing.

      Left turn crashes are the most common. Drivers have to wait for a safe gap, vehicles may be hidden, drivers behind may be impatient, and pedestrians crossing may be hidden by A pillars. It's interesting that they haven't come out with a solution for those crashes yet. Shouldn't turns be prohibited where there is a walk signal?

  2. Katie,

    Thanks for the extremely balanced and well thought out post. I should point out that many of the studies you reference looked at all right turns on red (which are legal if you stop first) and not necessarily rolling right turns on red. Therefore, the number of accidents and injuries caused by rolling right turns are necessarily less than the numbers you found from those studies. Since red-light cameras cite for rolling right turns (and not legal right turns where the driver fails to yield after stopping), and since 75% of the citations in LA are given for this behavior, we would contend that the resources being expended on the program are being wasted by targeting a behavior which is common (hence high revenue) but don’t actually cause very many accidents (low safety impact).

    In contrast to the above studies, we were able to discern which right turn on red violations occurred during a rolling right turn by combining two data sets of the SWITRS database (I believe we are the only ones to do this to date). Combining the databases is a bit complicated but once we figured out how to do it, we were able to find the data we were looking for. And as you note, the results show so few collisions each year that they barely even register in the database. We do believe that there may be some locations with such a high pedestrian traffic pattern that it might make sense to prohibit a right on red, if that would have a significant impact on safety. Our data mining does show that the critical aspect of all this is whether a motorist yields appropriately, whether or not they actually come to a stop first before initiating their turn. Additional signage such as the “right turn yield to pedestrian” sign might also be useful in high foot traffic areas.

  3. As a pedestrian it is not safe, as most times the car dosent even stop and looks to the left, often spends the entire light blocking the crosswalk as the oncoming traffics is too heavy. Also it causes the driver to be impatient never wanting to stop. I've been hit 5 times as a pedestrian crossing with this stupid law.. Lived in Europe and north America and people here in Canada feel like they have the right to never stop... Montreal banned the rtorl and it's much safer. Also I've seen people in wheelchairs unable to cross as a car trying to turn right blocks the crossing...

  4. I almost hit two pedestrians this morning turning right on red, and I am one of these people who does my best to watch for pedestrians and always yield to them when I'm supposed to etc. i myself am a pedestrian often times. what happened was I came to the light I always come to in the morning. Stopped right before the crosswalk, looked left to watch the traffic. It took about 10 seconds for a lane to open up but here's the thing... This is a multi-lane highway I was turning onto. As I turn right, I like to hold my gauze just a little bit on the other cars in the other lanes to make sure they're not changing lanes into the lane I'm turning into. So as I go forward I'm keeping my head left still for a split second. Well as I go to turn right this morning as I bring my head around convinced no one is coming Into the lane , all of a sudden there's two people one to two feet from my car. I slammed on the breaks but in very lucky they weren't right in front of my car or else I would have hit them. They hadn't quite made it that far yet. I admit it was a mistake on my part. I will make sure to look right to see if anyone had happened to appear in thec crosswalk before going. Ironically it wouldn't have been an issue if I had just pulled up over the crosswalk and forced them to go behind my car, but there wasn't quite enough space for that like there is in some roads. My car would have blocked some of the crosswalk. I now see that blocking right and red makes sense.

    1. This is one of those reasons that multi-lane crossings are especially dangerous for pedestrians. Even when drivers are trying hard to keep everyone safe, there is a lot of information to process at once. This could be a situation where a bulb-out or similar treatment that puts pedestrians further into the line of sight might be helpful.

  5. Drivers who encounter pedestrians in the crosswalk when turning right on red are obliged to yield to pedestrians who are crossing against the light. They shouldn't be there. This doesn't mean that drivers could justifiably cut them off. But let's compare apples to apples.

    1. The irony here is that we have a law which clearly favors motorists over pedestrians in that the motorist has legal permission to go through a red light yet pedestrians can still be ticketed for crossing on a "do not walk" sign. Apples to apples says let the pedestrians cross if it's clear.

  6. One large factor I think this article fails to acknowledge is that this is written from the assumption of motorist supremacy. Here's are two examples:

    - If an intersection is dangerous enough to warrant a signalized control then surely that level of control holds for all situations - turning right, turning left, going straight etc. The converse of this leads the motorist to think that surely they don't really need to stop.

    - There's also the sense that "collisions are less severe" or less frequent or less whatever which makes RTOR acceptable. This can be translated as "of course some pedestrians and cyclists are going to get hurt. That's the price we have to pay as a society for our motorized mobility." That's a load of BS.

    Finally it's worth mentioning that RTOR is almost universally banned in Europe, where they tend to be somewhat ahead of the game. Not always and not every where, but generally.

  7. There is an intersection in front of the city government building in my state that is a "No Turn on Red" intersection, and drivers routinely make rolling right turns on red. What is especially frustrating is that some of the most flagrant violators are law enforcement vehicles.

    I have repeatedly used the local police department website to request traffic enforcement at this intersection, and received no response.

    And people wonder why I wear an unfashionable reflective vest while commuting to work and school by bike, why I have an air horn on the bike, and why my bike has disc brakes.