Response to LTDA: Analysis of signal compliance, by mode

I thought I’d do a quick rough-and-ready analysis of the Licensed Taxi Driver’s Association videos that are doing the rounds today, which purport to show that people on bikes are serial lawbreakers.

Their analysis is based on two separate hour-long videos, one filmed in Camden, the other in Hackney. I obviously haven’t had time to sit through two hours of video, so I’ve just focused on the first fifteen minutes of the Hackney one. I will extend the analysis if anyone wants me to – the results probably won’t be very different, although the sample sizes will be larger, and make the result more statistically significant. Patterns of motor vehicle compliance will emerge which haven’t done so in this instance due to the small sample size.

My method is a direct copy of Chester Cycling’s own analysis of junctions in Manchester last year; go to his excellent post if you want more detail on the approach taken.

Here are my results, from the first fifteen minutes of the LTDA video.

Figures for signal compliance by mode, including opportunity to progress

Figures for signal compliance by mode, including opportunity to progress

Clearly, the picture is very different once we start to look at compliance relative to the opportunity to commit an infraction. Huge numbers of people on bikes had the ability or option to jump lights once they arrived at them – about 36% of all the cyclists in the first fifteen minutes of the video. This is obviously not true for the drivers of motor vehicles, most of whom simply did not have the opportunity to commit an infraction, because they are in a queue. Only 5 car drivers, for instance, had an opportunity to drive through the junction on a red signal, so it is not surprising that none did. This is a very small sample size.

It’s also worth noting that bicycles are, by far, the majority vehicle on the road at this particular junction, around 62% of all vehicles in this fifteen minute section of the video. Inevitably law-breaking is going to be more obvious when you are the majority road user.

When you also consider that signals are only necessary for motor traffic – pedestrians and cyclists are quite happy mingling through junctions without traffic signals – the fact that so many people, either on foot or on bike, are held at signals for the benefit of the minority road users is really quite unjust.

However, if we are looking simply at stop line/ASL compliance, cyclists are actually the best behaved out of all groups, expressed as a percentage of opportunity. Just 26% of those who had the opportunity to cross the stop line did so; there was far worse compliance with the ASL line by motorists, particularly motorcycles, 83% of whom entered the ASL when they could do so. Car drivers, HGVs and vans fared no better.

The bad news is that around 20% of those cyclists who had the opportunity to progress through the junction did so – this is the ‘full’ red light jump. I’ll leave you to watch the video to see how hazardous this is at this particular location. But based on this data, motorcyclists are actually worse – a third progressed through the junction with the signals on red (with the caveat, again, that this is a very small sample size).

Red light jumping also seemed to be a ‘copycat’ behaviour; if the ASL already had 3 or 4 people stopped in it, then it seemed to be the case that most other cyclists would stop. Conversely, slipping ahead through the junction seemed to occur more when others were already doing it.

Finally, this analysis does not include ‘amber gambling’, which seemed to occur on nearly every phase of the first fifteen minutes, irrespective of mode – this is something drivers and cyclists indulge in alike. There is a particularly bad example by an ambulance driver at 11:55 – this does not show up in this table, unlike the far less hazardous creeping across the junction by cyclists under a red signal. Only pure red-light jumping is included here, regardless of the actual potential for harm.

Make of this what you will – I thought I’d just add the broader statistical picture to the silly headlines.

This entry was posted in ASL Abuse, Stereotyping. Bookmark the permalink.

26 Responses to Response to LTDA: Analysis of signal compliance, by mode

  1. Chris says:

    I think you’re being a bit generous to the very first vehicle to say that wasn’t a full red light jump!

  2. Nick says:

    if cyclists are now the majority vehicle at this junction perhaps it is time to revisit the need for lights which were presumably installed to control motor traffic. Removal of the lights may not be the answer but the current arrangement may not be the best option either.

  3. Tom Chance says:

    Thanks for this, interesting stuff.

    I’d like to pick up on one comment, though: “pedestrians and cyclists are quite happy mingling through junctions without traffic signals”. That may be the case where a junction has been designed with that in mind, and where everyone understands the design and approaches it with due care and courtesy.

    But if pedestrians cross a road at a green light expecting to have a clear use of the crossing, that really isn’t true. There is a junction in Peckham I use daily where, daily, I see people on bikes anti-socially wiggling through pedestrians crossing at the lights. The pedestrians there definitely aren’t happy, nor are the many who complain about bad behaviour elsewhere in London. The same can be said of pedestrians walking out into the road into cycle lanes without looking. Sure, a cyclist can in theory brake and mingle along, but it isn’t a happy experience.

    Sometimes road behaviour is about courtesy, and street design about making life pleasant.

    • Thanks Tom, completely agree. Sorry if I wasn’t clear – my point was a more general one about how traffic signals have basically arisen as a response to motor traffic. In the most walkable and cycleable cities, they are far less numerous, because junctions don’t need to be signalised (obviously there will still need to be priority rules, and things like zebras, where appropriate).

    • Tim says:

      Just to add to aseasyasriding’s reply, I think a good example of this which has been mentioned before before is the green scramble:

      Of course lights make sense at pedestrian crossings, to allow pedestrians to cross safely where there are large numbers of motor vehicles or cyclists. But when it comes to junction management, the green scramble video makes it clear that the lights are there for the cars, not the bikes.

  4. On third phase blatant RLJ by lorry travelling in opposite direction (which is more dangerous, because it bisects trunk of T).

  5. Phil C says:

    Having not watched the video, I am making an assumption here based upon my own observations (a method seemingly very popular -and, if you are an association, official sounding)

    Taxi & minicab drivers in my area (Manchester) appear to me to be one of the worst offending groups to transgress the ASL, often leaving cyclists nowhere to go. I refuse to believe that driving (or cycling) cultures are vastly different from north to south of the country. Therefore the only explanation I can give for NO transgressions by taxis in this video is that they had prior knowledge it being filmed and were asked/warned not to stop past the line at these particular junctions.

    I hope the association who commissioned the videos didn’t inadvertently let it slip where they were filming and unintentionally cause the figures to be skewed…….. Now, Who was it that commissioned the videos again…….?

    • zero (the unsignified) says:

      You might refuse to believe it, but as someone who has worked as a cycle courier in both London (13 years) and Manchester (12 years and counting) I’m afraid I’m going to have to tell you that drivers in the North are, on the whole, much worse than those in the South – excepting taxi drivers, ironically.

    • Edwin says:

      It’s a film for the London Taxi Drivers’ Association, so I’d have thought it unlikely they would include any transgressions by taxis. Besides, there are only 5 taxis in the films, and there is really not enough data here to draw any conclusions at all.

      “I will extend the analysis if anyone wants me to – the results probably won’t be very different, although the sample sizes will be larger, and make the result more statistically significant.”

      I don’t agree with this — the results can easily be completely different when taken over a larger sample size (e.g. more than 5 taxis), that’s the whole reason for requiring statistically significant results. To get anything meaningful that we can generalise from, we’d also need a representative sample of different junction types and different times of day.

      If the goal of the analysis is to determine whether cyclists break the law more than other road users, then it’s also pointless to look only at traffic light compliance, and not at parking, dodgy overtaking, etc… Besides, as explained above, there are mitigating reasons for some of the non-compliance by cyclists. So, really, a statistical analysis of this dataset tells us very little. I think it’s better to spend time promoting practical improvements rather than engaging with these “debates” at all.

      • I agree that a broader picture of law-breaking across London, at multiple junction types, at different times of day, would be much more relevant – to say nothing about the relative potential for harm posed by different forms of law-breaking.

        My point here is only really to demonstrate that even the LTDA’s attempt to prove something with a desperately shallow analysis fails on its own terms.

  6. pm says:

    I think it would be very easy, if one cherry-picked the junction, to produce a video showing motorist red-light-jumping in equally high proportions.

    As you point out, only the cars approaching the junction as the light changes have the chance to jump the red or not – once one has stopped there are no more cars to include in the sample till the next phase. And to match the taxi-driver study methodology you have to discount all vehicles going through on green.

    The key issue is the universal tendency of drivers to accelerate if the light goes amber as they approach. Not only is going through on amber itself often a form of rule-breaking (I think you are only supposed to do it if it would be dangerous not to, right?), but it also leads to at least one car on each phase – often, in my experience more than one – not quite making it and hence going through on red. On some junctions at some times the number of vehicles doing this outnumber those stopping correctly by a large ratio. And that’s not even considering the ASL intrusions.

    • It’s actually quite clear that the LTDA have cherry-picked these junctions, as ones where RLJing by bikes is much more likely. They’re both effectively T-junctions, where proceeding straight on by bike will only bring you into conflict with pedestrians using the crossing directly in front of the stop line.

  7. cyclestrian says:

    So what? I could pitch up with video camera plus speed gun on almost any residential street and demonstrate that people **in cars** are speeding. Evil drivers. Or spend an hour or two in London and show LTDA members breaching 30, stopping where they’re not supposed to and RLJing.

  8. Blaise says:

    Red lights act as a subsidy for motor vehicles that can accelerate quickly to top speed. For cyclists, it’s more costly to stop and accelerate all the times. It’s more efficient to have some momentum. A city with a lot of red lights (like London) is clearly a city designed for motor vehicles. As you say, fewer red lights would not be necessary if there were fewer motor vehicles and more bicycles.

    The reality is as well that, as a cyclist, the cost of jumping a red light is low. To my opinion, most on the red light by cyclists in this video are reasonably safe. This is a T intersection and it’s easy to see cars coming from the right (I imagine that the camera was deliberately positioned here to catch a lot of cyclists). Moreover, it’s true that you’re less likely to be fined if you jump a red light as a cyclist. And I would add that one of the reason why cyclists jump red lights sometimes is to get ahead of cars that may block them if they turn or try to overtake them at a narrow junction. The cost-benefit analysis of jumping red lights is very different for cars and bicycles.

    Car drivers are very moralistic with cyclists who jump red lights because their reference is driving a car and jumping a red light when you drive a car is seen as one of the worst offence. For bicycles, I don’t think it’s the worst offence and I rarely see cyclists jumping red lights in a way that put them or others in danger. I have less sympathy for cyclists who do not use lights at night or the ones who ride on sidewalks. From a cyclist point of view, the worst offences that cars regularly do are overtaking me too closely (very annoying but not so dangerous in fact) and advancing in the ASL (and parking on the cycling lane and forcing their way and driving too close and not use turning lights etc.). Drivers don’t see these two offences as badly as jumping red lights but they are more dangerous for cyclists.

  9. Pingback: Blood on the streets - Page 16 - London Fixed-gear and Single-speed

  10. In my experience of commuting in London, motorists run red lights at every junction, every cycle of lights. By ‘run’ I mean continue through the red light at the end of the cycle. In doing that, they risk hitting cyclists who, by and large, only ‘jump’ red lights (proceed before they turn green). I’ve seen some near misses of red light jumping cyclists with red light running motorists. So in my view we need to raise this issue of how prevalent and dangerous (because at speed) RLR is with motorists.

  11. Mike Chalkley says:

    2 points:
    1. Red light jumping by cyclists endangers the cyclist – not motorists. The law is supposed to be there to protect us from the actions of OTHERS – not ourselves.
    2. “Amber gambling” is in fact as serious as RLJ in the eyes of the law. If the lights turn amber you MUST stop unless it isn’t safe to do so.

    • Blaise says:

      The law is supposed to protect yourself too. You must wear a seatbelt, motorcyclists must wear a helmet, it’s illegal to take drugs (with few exceptions such as alcohol) etc. And I would argue that jumping a red light can endange motorists too if the cyclists provokes an accident. But I still think that jumping red light for cysclists is not so dangerous because the majority of them jumps red lights when there are no cars approaching.

    • Tim says:

      > “…unless it isn’t safe to do so…”

      But of course that’s the tricky bit isn’t it. Who gets to make that call? If you’re driving like a wally in the first place (ie fast) then it’s likely to make a stop more abrupt and dangerous.

  12. The number of cyclists failing to stop on red increases as the video goes on, so if you’ve only analysed the first 15 minutes, you get a different view than if you sit through the whole hour.

    LTDA’s claims are still well off, but I make the rate of infraction-per-opportunity over the whole hour at Hackney more like 46 percent, dropping to 32 percent if you include riders coming in from Queensbridge Road, all almost all of whom stop at the light.

  13. Tim says:

    I have to wonder, how come the massive discrepancy between the 20% (for cyclists) on this page and over 50% (in Hackney) mentioned in the Standard article? Are they including “line-passers” ? If so it’s surely offensive not to report the same for drivers (for balance).

    Having watched the same 15 minutes myself I didn’t see a single incident where I felt anyone was reckless – ie where anyone was put in danger. Admittedly there aren’t many pedestrians crossing there and you couldn’t see the other side of the function too well (where there seemed to be more pedestrians). And of course this isn’t to say that it’s always OK to jump reds, but it really isn’t the same as drivers doing it.

  14. Pingback: What traffic policing COULD be like | Road Danger Reduction Forum

  15. pm says:

    Your own video seems relevant here…

    The publicity this LTDA video has received is quite a contrast with how little anyone even notices the habitual red light jumping behaviour of motorists as shown in the video on that post.

    The simple reality is, most road-users only obey the rules insofar as breaking them would either (a) carry a serious risk of being caught and fined or (b) endanger themselves or their vehicle. Concern for the safety or convenience of others or for the general principle of the ‘rule of law’ really comes a long way behind those two for most.

    Cyclist and motorist violations both follow that simple principle, but because motorists have number plates they perhaps worry about (a) a bit more, but because they are less vulnerable they worry about (b) a lot less.

  16. Simon Smith says:

    good work! I watched the first 30 mins and the numbers were just not there? So I counted for first 10 minutes, 62 affected by red light, of which 12 proceeded through, 1 in 5, still not good but way off 53%….see they have started a petition for cycle licenses. I was surprised just how many riders there are, amazing! It must be a nightmare for taxi drivers, so they want cyclists off the roads!

    I suggest a nice group bike crawl around London, obviously stopping at all the red lights, just to show how well behaved we are!

  17. James says:

    Amber gambling is the thing that concerns me. As a cyclist approaching an amber light I’ve often slowed down only for the car behind me to assume that I was going to go through the lights and they accelerate and only just manage to squeeze pass me as I stop in the box. Some cars almost drive you through the red themselves.
    Drivers have a hierarchy of laws of the road and the ones at the top are driving through red lights and driving on the pavement, things that they would never do but often see cyclists do. Rules or policies regarding speeding, using a mobile phone, turning without indicating or stopping in a box-junction are lower down this hierarchy and are therefore optional and we see car drivers doing this all the time but not so much from cyclists.
    Incidentally, we did see many cyclists going through the lights but not single issue occurred because they used their common sense and merged. This also happens in the US where drivers are trusted to drive through red lights in certain circumstances. Drivers just don’t like it because it is the one law they choose to obey and cyclists “get away with it”. As motorists are clearly the minority at this junction they should be thankful that they have got so much of the infrastructure for their mode of transport. Let’s point the camera the other way and see how many were on phones or approached at speed or overtook a cyclist with less than a meter of space.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.