Penalty for three second stop

Of course it will be changed. Imagine a queue of stationary traffic at a red light. The light changes to green. The first car moves off, but the second car would only be able to move 1 car length and then stop again to wait for the first car to clear the junction. Repeat for every car in the queue and you'll have a huge slow-down in traffic. Even in free-flowing traffic cars crossing a junction will seldom be as much as a junction-width apart.

In practise it is *usually* posssible to see that there is sufficient room after the junction to accomodate at least 2 cars, so you can safely follow immediately without risk of there not being space to clear the junction.

In the OP's case however, a large vehicle blocked his view of the traffic situation on the other side of the junction. He followed immediately after the vehicle in front, wrongly assuming that there would be sufficient room to clear the junction.

No it won't because they'll all have to stop behind the staionary cars after the junction anyway, you've just added a gap the width of the junction.

That's the only time it could slow it down, but that's when it's not an issue.

I'm afraid I agree with Cynic's assessment, and in any event there is concrete evidence that strict enforcement of box-junctions can often lead to reduced throughput.

e:

As Cynic describes, it is essentially "the fear of crossing the junction until the way is visibly clear" - as opposed to the usual standard of driving which is "crossing the junction if it seems likely that the way will be clear once the junction has been crossed".

At a light controlled junction, the failure to cross the white line until the way is visibly clear can mean potentially a whole junction's- width of queueing vehicles remain behind the line when the light turns red, whereas if they had crossed on the *expectation* that their way would become clear, then many more cars would have crossed the junction before the light turned red. Also, the increased number of cars queueing (as a result of reduced throughput) and reduced queueing space (because the junction itself is now kept totally clear) can cause more congestion upstream.

I expect a lot of these problems can be ironed out with very fine- tuned signal timings (particularly if the previous timings catered to lax enforcement of the box junction), but certainly TfL found that some junctions suffered due to enforcement:

While enforced box junctions may slow traffic down, there seems to be a misconception here that the rules of the road are just about maximising the flow of traffic. It seems to me that there are a good few cases where the opposite is true. Speed limits are an obvious one.

You are incorrect. If everyone temporarily paused at a junction for the time needed for the car ahead to cross the junction from a standstill, it would greatly reduce the number of cars that cross the junction in each light cycle. That would in turn *greatly* decrease the carrying capacity of the junction. It may make little difference in light traffic flow, but it will lead to traffic jams occuring at a far lower traffic density than when drivers follow with a closer spacing.

Despite your previous profession you appear not to understand traffic flow. You may like to reflect on the fact that closing one lane of a two-lane road would at most slow traffic to half its previous speed if using simple analysis. In fact, if the traffic is above a certain density threshold, it will cause a huge traffic jam where miles of cars are at a virtual standstill. Which I am sure most posters will have experienced.

Not if the cars on the oither side of the junction are *not* stationary.

I thought that it was being advocated that you should stop and wait for the car ahead to clear the junction *all* the time. In a traffic-jam situation I agree that that is what you should do.

That's not concrete evidence; that's a theory that might predict reduced throughput in some circumstances. Other theories would predict increased throughput; we've all seen junctions blocked by inconsiderate drivers queuing in the box.

That's not "certainly", it's a very brief report of preliminary findings of a small study (only 16 boxes) back in 2006. Before making a judgment one would want to know the actual figures and methodology, and the follow-up results on more boxes.

Indeed.

I don't know whether you're confused, but no one here is saying that enforcement of box junctions *necessarily* leads to reduced throughput overall. What we're saying is that the circumstances can be such where additional enforcement does lead to an overall reduced traffic flow, and I've given you evidence of TfL's experience of that - I really don't see what additional evidence needs to be provided.

No, I don't think so. Just statistically literate.

I know, and that's what you need to prove.

What was the statistical power of the TfL study? What was the p-value for their conclusion of "up to 150 fewer vehicles an hour passed through"? What does "up to" mean in this context? Two?

You *never* draw firm conclusions from a report like that one in The Times. It's worth knowing about, but that's all.

Well, if you're an expert on these matters Les, then I bow to your superior knowledge. I'm merely supporting Cynic's contentions that enforcement has, in the past, reduced throughput or even made junctions untransversible without breaking the law, and that was really all Cynic and I were saying.