Traffic Light RedYou may have heard from this Herald article that AT has just released some stats about red-light running by drivers and cyclists. And yep, at first glance the stats look bad.

Cyclists red-light running clearly outnumber motorists and cyclists are clearly over-represented due to them being less numerous, as AT correctly states (and we don’t want EITHER cyclists or motorists to run the red light!).

But a lot of context is also missing:

  • Three of the four locations were on Quay Street and Tamaki Drive. Why is this important? Because according to Auckland Transport’s own statistics, we have 1,346 cyclists there daily (March 2013 AT survey, Tamaki Drive / The Strand) making this one of the busiest cycle routes in NZ! Therefore, you are likely going to get higher red-light running totals too.
  • In fact, the surveys were undertaken over TWO days (6 peak hours each day). So we really need to double the daily cyclist counts for a fair comparison. As an example, at Solent Street – just east of the The Strand counting location – within 2 days AT counted 48 cyclists red-light running. 48 cyclists breaking the law compared to probably well over 2,000 cyclists using the route (due to different count times, we can’t make an exact comparison). This changes the context considerably.
  • Out of the four locations, three had very high cyclist red-light running counts. However, one location (Newton Road / Ponsonby Road) had 39 cars run the red light within two days while only 14 cyclists did so. And that intersection still has a very busy daily cyclist count of 858 (March 2013 AT survey). Why is this important? Because it shows that the problem is not automatically the same everywhere and that the study may have chosen locations that were unrepresentative of the overall picture.
  • Lastly, the survey seems slanted towards intersections KNOWN to have significant cyclist red-light running, likely because AT commissioned the study while preparing a related education campaign. That is fine with us – you need to study what you want to improve. However, releasing it without such qualification, in the context of this latest tragic accident… It is a bit like publishing a study into the the “Top 10” worst speeding spots in Auckland under the headline of “Auckland Speed Survey” – with many people on the outside given the implication that all car drivers speed like that…
A safe cycling intersection - how many lives would be saved each year?
A safe cycling intersection – how many lives would be saved each year?

As has been mentioned in past posts, a paper and related presentation by AT’s own Daniel Newcombe shows cyclists red light running as about the same as that of pedestrians (~4%), while a Melbourne study showed it at 7%. A very different picture from that implied by AT in this narrower survey.

As CAA has said before – we acknowledge red light running by cyclists as both an actual safety issue, and a perception / public image issue. But a bit more context would have been appreciated for an informed discussion.

Let’s have a debate based on facts, not half understood myths and anecdotal “evidence”.

General News
Share this

37 responses to “The problems with the red-light running stats from Auckland Transport

  1. Why give daily cycle counts when the survey was only taken over a few hours each day? I thought CAA have access to the cycle counter data so a more accurate number could be found by looking directly at the database.

      1. My complaint was that you reported the AT total counts as daily counts. Looking them up now, I see AT was actually only counting there during peak hours, as they were with the red-light runners.

        So at Tamaki Dr/The Strand we had 507 cyclists in the morning and 492 in the evening. Doubling that gives the 2000 for the peak hours over two days.

        That gives about 2.5% of cyclists at that intersection jumping the red, less than expected based on the other estimates you quote.

        1. David, I am still not understanding why you feel the need to criticise us for providing data that AT didn’t even seem fit to provide at all. Especially when we acknowledged the issues with comparing them AND you came to a similar conclusion.

          Also, the AT red light survey noted that it was taken from 6am -9am and 3pm – 6pm on both days, while the cycle counts were taken from 6:30am and 9:00am and between 4:00pm and 7:00pm. So they were missing half an hour in the morning, and counting one hour off-set in the evening.

          Exact comparisons can thus not be made easily, and thus we used an approximate, and conservative “well over 2000” number for the cycle number that went past during the two day’s red light survey period, which acknowledged that you couldn’t simply multiply 1,346 by two and call it a day.

          1. I was just trying to understand what you were doing — I only came to the same conclusion once you clarified it. I’d say the peak hour comparisons are much closer than the dodgy “daily” figure that AT produces.

            Would be nice to see a link to the AT red light release if you have it.

          2. Hi David, sorry, no link. AT just sent it to a couple of people (not including us, but…)

        2. “My complaint was that you reported the AT total counts as daily counts.”

          Yes, we noted these as “daily cyclists”. Which is exactly what the AADT (annual average daily traffic) given in the study means. You will find that this is perfectly normal, and used in assessing car traffic in the same fashion.

        3. In a sense of exactness, the 2000 also includes those who arrived while the light was green anyway, so your exact percentage of “how many obeyed a red light vs how many did not” really depends on another set of data that is totally non-existent for this one (we don’t know the number of users who arrived while light was red). So maybe the actual number of rule-breakers may be 4% or 6% or whatever, which again would fit closer to Daniel’s study and the Melbourne study. We don’t know. All we know is that it is a small fraction of the total, unlike it is being portrayed.

          1. Good point. Isn’t the 4% number that Daniel produces of the same type as the 48/2000 you reached above?

    1. Well, look at this. If I am reading the responses from the automatic counters right, then the counters were broken / off for 75% of the time during those 2 days.

      Monday, the eastbound counters seems to have been both off.

      Tuesday, the eastbound counters seems to have been faulty – showing only 4 cyclists total.

      Monday, the westbound road counter seems to have been OK, counting some 400-odd users. Tuesday, exactly from midnight, it seems to have been off, zero users all day counted. Same for off-road path, with some 60 users on Monday, and zero users on Tuesday.

      Which is why it is good to have that manual count, and so fittingly close in time and location too…

  2. What also needs to be considered is the type of intersection. If it is a T intersection, then I do know a lot of cyclists who will ride through it, if they are riding at the top of the T. Primarily so that they are out of the way of following traffic. Not unless the trafic perfers to build up behind them.
    But there does seem to be an issue with data sets taken. If AT is going to look at specific intersections and then allow everyone to extrapolate that data, then maybe AT would like to monitor the Shakespeare Road cycle lane at the Tahoroto road intersection. On any given day, AT would be able to see without fail almost every vehicle driven down the cyle lane. Extrapolate that, and you would then say every motorist in Auckland uses the cycle lanes as an extra lane.

  3. Does anyone know why AT can one month publish an award-winning, balanced, professional paper that references credible local quantitative and qualitative data and appropriate international experience and relates it to Auckland’s current situation.. in a way that is accessible, persuasive and helpful to all….

    …and then another month publish such a load of rubbish?

    Is it supposed to achieve some kind of “balance”?

    One of of organisational flip-flopping is that it is an early indicator of their demise. How long shall we give AT in its present form? 3-5 years?


    Sorry for the caps but it needs to be said loud and clear that their is NO EXCUSE EVER for running a red light. Until the cyclist lobby takes this seriously they will never get the support of drivers.

    1. But motorists do it all the time. And how can cyclists park in the green box at the intersection, when motorists purposely park there? Personally, I am not trying to justify it. I am understanding why, so that appropriate measures can be taken to remediate it, or stop it. Are cycling groups doing it? Focus on them and get their organisers to stop it, or prosecute them. Is it so that the cyclist doesn’t get run over by motorists keen to get through the lights before they go red again? There are intersections where Auckland bus drivers will purposely veer inn over cyclists when the lights go green, even if the cyclist was at the lights first.

      1. “But motorists do it all the time”

        This is kind of saying because others do it, so I can do it as well. With this kind of attitude, cyclists are not going to gain respect and support from other road users I’m afraid. How about changing it to “I know others do it, but I will not do it”.

        1. And why do you keep implying that we do? Repeat after me: Almost 19 out of 20 cyclists already don’t do it. BOTH recent AT statistics AND the Melbourne survey show this. All you do is perpetuating stereotypes, and then asking “the cyclists” to show repentance for them.

          1. How does your “19 out of 20” figure match with the survey results of Daniel Newcombe in which much higher percentages of cyclists admit to running red lights under certain circumstances? For example, he says that his survey had only two thirds always obeying the red on a left turn. (I’m not sure about the provenance of this survey, it may well be self-selected and unreliable).

            It looks to me to be more accurate to say that 19 times out of 20, cyclists obey the red light.

            The distinction is important: saying 1/20 break the law sounds like it is a marginal group ruining for the rest. 1/3 occasionally running a red makes it clear a significant minority of cyclists see red-light running as acceptable in some circumstances. Obviously, they disagree where to draw the line.

            Just to be clear, I’m not knocking your approach here, you and CAA are doing a great job on this. Greatly appreciated.

          2. We said “almost”, because some surveys (like Melbourne) say 7%, so we couldn’t say “19 out of 20”.

            We say “almost 19 out of 20” because Daniel’s survey showed that at NON-Barne’s dances (and we have how many of those in Auckland? 20 out of 2000 signals?) the red-light running of cyclists seems to be around 4%, and we say “almost 19 out of 20” because the AT survey splashed across the Herald showed that at Solent Street, the amount of cyclists running red lights (as a percentage of ALL cyclists) is around 2-3% of the total cyclists going through (we don’t know exact number of cyclists who were supposed to stop). All these show that “almost 19 out of 20” seems a fair approximation, especially when many other participants in the discussion give no estimates or counts at all, or spout stuff like “almost all” or “90% of cyclists run the red light”.

            Saying “almost 19 out of 20” is thus reasonably well-supported by stats.

            Would CAA prefer a more comprehensive, more Auckland-wide, as-neutral-as-we-can-get survey of this? Heck, yes!

          3. Fair enough.

            I still think it is more accurate to say “19 out of 20 times a cyclist will obey a red light” rather than “19 out of 20 cyclists obey red lights”.

        2. Respect? Respect? Okay, so you are saying it is fine for motorists to drive along and purposely punch cyclists, just because cyclists don’t deserve respect. This is the result of these misleading statistics. Last weekend two cyclists were physically assaulted for no other reason than that they were cycling. In one instance, the motorist intentionally stopped, and waited for the cyclist. You are correct that respect must be earned, and so far, many motorists do not deserve any respect at all.

          1. I never said it is fine for anyone to purposefully assault or attack other road users, not for motorists, not for cyclists, not for pedestrians.

            You are saying motorists are generalising about cyclists, but cyclists are the same. I know a lot of motorists don’t obey road rules and some purposefully abuse other road users. I make sure I don’t do that (and I never did) and I try to educate others to maintain good standards if that’s achievable within my ability. Correspondingly, I would expect the same from cyclists, but when approaching two male cyclists at my work, I was told “it’s stupid to follow the road rules; those are not for cyclists” and apparently this is not uncommon among cyclists.

          2. Road Rage is not confined to any particular mode of transport. I have seen car drivers yell abuse at cyclists but equally I have seen cyclists wave fists and scream abuse at cars and pedestrians. You get inconsiderate people everywhere. This is still no excuse for the vast number of cyclists refusing to obey laws or picking and choosing which laws they obey.

          3. There is a slight difference between yelling, and fist waving, and being physically attacked for no reason at all, apart from riding a bike. The events you talk of phil are the result of some interaction beforehand. The one I speak of was in effect a hunting party. Big difference.

    2. It may seem strange but sometimes you have to run red lights to get through a intersection. The road sensors do not pick up bikes – I ride through a number of lights that you never get a green light unless a car is going the same way – what would you do after 3 complete phases of lights with no green light to proceed. One of these sets of lights I sit in the middle of the road trying to turn right on a green that never comes with cars passing you (usually above 50km per hour) on the left – it is a scary place to be – I would sooner run a red 1. because you know its not going to go green and 2. you are out of harms way – all about being safe.

      1. Fair enough; I appreciate sometimes there are technical difficulties faced by cyclists. However, I would use pedestrian lights and crossings (push through not ride through) instead as I think this would be the safest approach for myself. Yes, it will take longer, but slow is better than sorry.

        1. Problem is though observer, is that on right hand turn signals, by the time you are caught by a ligh, you are stuck in the middle of the road. To use the pedestrian crossing means walking to the left hand side of the road, crossing lanes of traffic, to then cross to the right. Incidentally, crossing at the lights without a signal, as a pedestrian, is also illegal.

          1. Maybe for others, but not necessarily a problem for me if I was aiming to use the pedestrian crossing in the first place (so I wouldn’t be stuck in the middle of the road). I just think riding along with cars at an intersection when turning is a bit dangerous (for me anyway) as cars go a lot faster and it may not be easy for every car in the queue to see me. In addition, the condition of the road I was going to turn into may not be clear until I got there (for example, there might be a car trying to parallel park, etc.). But that’s just me.

          2. Observer, the problem is when there are no vehicles turning right. And if one cycles on roads that don’t have a lot of traffic, or at times when the traffic is minimal, one is stuck. There is a bike route I do, which has 80 sets of lights on it. 40 out, 40 back. That is a hell of a lot of lights to get off the bike and walk through, if one is assuming that they all won’t pick up bikes. And for right hand turns, that means walking through two sets of phases. One to cross the road, and then another one to get to the left hand side of the lane.

          3. I was only talking about right-turning ones raised by Brian, not the straight-through ones. Yes, using pedestrian crossings will mean going through two sets of lights. Personally, I would rather do this than running the risk of getting stuck in the middle of the road. This would be my decision for my very own safety and it also means I would not have to break the law (running a red light). If I wanted to go somewhere quickly, then I would choose a different mode of transport. Where speed is not important for my journey, then safety is certainly the top priority for me (though not saying I would drive unsafely).

            P.S. I rode a bike on a daily basis for 7 years in Shanghai, a city of 20+ million people where there were far more private cars, trucks, buses, taxis, bikes and pedestrians, and not all roads had dedicated cycle lanes. Didn’t think cycling was an issue back there. I still struggle to understand why it is such a contentious issue here in NZ.

          4. “I still struggle to understand why it is such a contentious issue here in NZ.”

            Observer – I agree. Why so many people are so determined to attack cycling and assign it to some kind of lunatic fringe is beyond all of us at CAA. We constantly try and educate AT/NZTA on the benefits financially/health/urban planning wise to Auckland but it mostly falls on deaf ears.

            As you say, in many countries it is a normal part of life and not that big a deal. Just as it was in NZ for many years until we turned everything over to the roading lobby.

            The reason is that NZ is one of the most auto dependent countries in the world and NZers have been carefully indoctrinated to believe that anything other than the private car is some kind of socialist experiment to take away basic freedoms. There is nothing special about Auckland/NZ that should mean cycling cant work here as a means of transport – we just need some political will to get things going. Then we will stop falling even further behind the rest of the developed world.

        2. The particular intersection I am talking about is turning right from Richardson Rd in to May Rd – to get at the pedestrian button you would have to ride on the footpath as the council have put up a fence to separate the footpath from the road due to there being a school on the corner – also the crossing takes you to the wrong side of May Rd, you would then have to the second pedestrian crossing to get on the right side of May Rd. The whole process would be more dangerous than going through a red light when the road is clear in all directions.

    3. Actually there are very good reasons to run a red light. Or more accurately to ride carefully through a red light. Many other countries recognise that in many cases it is safer for the cyclist and for motor traffic for the cyclist to be out of the way during the normal green light.
      If the surveys are to be believe most of the “red light running” by cyclists in Auckland is exactly that, cyclists safely getting out of the intersection.
      Contrast that with cars, where most of the red light running is cars racing through a light that has just gone red, usually at high speed.

  5. Instead of spending time figuring out why AT’s survey shows a much higher red-running rate compared to the other studies, why don’t you guys make more effort in educating cyclists to obey the road rules? Red light running should not be jusified in any case. If cyclists want to share the road, then please obey the rules – this should be a minimum requirement. The traffic lights are put there for a reason. Until the time there are separate lights for cyclists, everyone should react to the lights that are put there for all. Road being quiet or it being a left turn – these are all excuses really.

    By the way, AT did provide the survey report to CAA as reported by NZ Herald, so please make this record straight and don’t mislead your followers.

    1. “Instead of spending time figuring out why AT’s survey shows a much higher red-running rate compared to the other studies, why don’t you guys make more effort in educating cyclists to obey the road rules?”

      This is so wrong. You are asking us NOT to figure out the “why”? You are simply asking us to take a misleading survey that threatens cyclists – because lots of agressive people on the road now have new ammunition without the context that the AT survey showed about 2.5% of all riders disobeying the rules?

      You are basically asking cyclists to say “mea culpa”, and for cycle advocacy groups to lead not the education effort (which we have done, on a shoestring budget, BTW, we are unpaid volunteers with full-time jobs and families) – you are asking us to lead a grovelling for forgiveness, as if we were all guilty. Forget it.

  6. Ignoring the impossible argument of which group follows the rules, because many cyclists drive and many drivers cycle, so really it comes down to the simple truth that people are morons.
    To “educate” cyclists not to run rd lights, as this “seems” to be the biggest complaint, as Max say’s why do they run them.
    I’d suggest the big reasons are (and possibly what made up the numbers), and suggested resolutions.

    Left hand turns. Well, turning left in front of any vehicle is stupid. Instead of lumping the cost onto trucks, why not utilise the green strip at the left of lights, and continue it around the corner, with a give way rule of turn left, ONLY if it is safe. And NOT through pedestrians.
    T Intersections. If travelling through the top of the T, why not continue the marked cycle lane. If there is room.
    The reasons for these two, is not to allow cyclists to get ahead of everyone else, but to keep them out of the way. Eliminate the need for motorists to try and push pass the cyclists at the lights.
    Peletons. This isn’t a hard one. I agree. Either make them cut down their size number, AND/OR put the message out to the group leaders that they need to follow the rules, which includes the red lights. Target the leaders and prosecute if need be. I suspect that much of the numbers were made up by these groups cycling along the waterfront.

    Catch the low hanging fruit first.

    1. I believe CAA has mentioned solutions like the ones you give for left-hand turns and T-intersections to AT but have been rebuffed.

      As for working with pelotons, see CAA’s Good Bunch campaign which has been rather successful.

Comments are closed.