Disclaimer: This is a post about a research project I led at work. My policy is not to blog about things that I’m working directly on, but in this case the research has already been reported elsewhere. All facts and figures in this post are drawn from a summary of my research and some related work that was presented to the City Centre Advisory Board.
Since I started working in the Auckland city centre in early 2012, I’ve noticed big increases in the number of people walking around the city centre. During the midday peaks, many footpaths are now filled to the point of overflowing. And there are increasing numbers of people walking around until late in the evening. This isn’t just a figment of my imagination: the available monitoring data confirms that the footpaths are getting busier.
This is good economically (more people walking around means more retail opportunities) and good socially (busy areas are generally safer and more interesting). So viva walking.
It’s also worth noting what hasn’t happened over the same time period. The number of cars entering the city centre during peak times has basically held flat since 2000, meaning that public transport has catered for all of the growth. This means, in turn, that the ratio of people on foot to people in cars has significantly increased.
There are a lot of places in the city centre where people in cars and people on foot must interact. Intersections, for instance: Signal timing dictates how much time cars must wait for a green light, and how much time pedestrians must wait for the green man.
The common (but not universal) practice in New Zealand is to make decisions about signal timing based mainly on minimising delay for vehicles. People on foot are accommodated, but they aren’t necessarily prioritised. In a lot of contexts, this isn’t totally crazy – there are a lot of cars and not a lot of pedestrians, so you’ll get an approximately optimal outcome that way.
But in a busy city centre environment where there is a high and rising ratio of pedestrians to cars, you have to ask whether this is the right way to make decisions. On Queen St, there are around four times as many pedestrians as vehicles. On High St, there are around 13 times as many pedestrians as vehicles.
That’s what we tried to do in this research. We found that people weren’t counting delays to pedestrians at busy city centre intersections. So we set out to count pedestrians and pedestrian delay at two city centre intersections – Queen St / Victoria St and Upper Queen St / Karangahape Rd.
The Heart of the City’s pedestrian counters provide a great source of data on trends in pedestrian volumes, but they’re not granular enough to know what’s going on at individual intersections. For instance, there are a lot of people walking along Queen St immediately north of Victoria St – an average of around 38,000 per day in 2016. But how many of those people cross the intersection when they arrive at it, and how many turn onto Victoria St?
So we went out and gathered some data on what was happening at these intersections. On a sunny February day during the busy pedestrian hours (noon-1pm), we set up a few video cameras to record how people were using these intersections – ie whether they were crossing the street or going in a different direction. Here’s what we found for the two streets:
We then estimated the average delay per pedestrian based on data on signal phasing, cross-checked against measured pedestrian behaviours. We assumed that:
- People who arrive during the traffic phase must wait until the next pedestrian phase
- People who arrive during the pedestrian phase with enough time to safely cross will simply walk across, with no delay
- People who arrive near the end of the pedestrian phase, with too little time to cross safely, must wait until the next pedestrian phase.
We found that:
- The average person walking through the Queen St / Victoria St intersection was delayed by 27 seconds relative to the ‘free flow’ walking conditions that they would experience in a shared space or pedestrianised environment
- The average person walking through the Upper Queen St / K Rd intersection was delayed by 38 seconds relative to ‘free flow’ conditions.
To conclude, we estimated annual pedestrian delay outcomes using the Heart of the City’s pedestrian count data and applied the NZ Transport Agency’s Economic Evaluation Manual parameters to estimate the equivalent monetary value of wasted pedestrian time. The figures added up quite rapidly:
- People walking through Queen St / Victoria St experience annual delays of 161,000 hours, which ‘costs’ around $2.2 million per annum
- People walking through Upper Queen St / K Rd experience annual delays of 40,000 hours, which ‘costs’ around $0.7 million per annum.
We also extrapolated our findings out across the entire Queen St corridor based on pedestrian count data and information on signal timing. The results were interesting – we found that delays at successive intersections, all with relatively high pedestrian volumes compared with car volumes, added up quite quickly – to perhaps $12 million in cost of delay per annum.
I’d like to make three points about our findings.
The first is that pedestrian delays in the city centre can have ramifications for Auckland’s transport system as a whole. The city centre is a major public transport destination, but it’s also an essential transfer point between services that serve different parts of the city. People transferring between services must often cross major intersections, where delays may cause them to miss a bus or frustrate them to the point of skipping PT and driving instead. (As highlighted in Shan’s great posts on Wynyard Quarter connections.)
But there’s also an opportunity. Building light rail down Queen St and creating a pedestrian mall up and down its length will significantly reduce delays for people walking on the corridor. Our analysis shows that reduced intersection delay for pedestrians could make a significant contribution to the benefit-cost ratio for the project.
Second, while it may be challenging to reduce pedestrian delays to zero, due to the need to let vehicles through many intersections, we can significantly reduce these delays with relatively simple changes.
For instance, we had information about recent changes to signal timing for the Queen St / Victoria St intersection. At the moment, there’s only one pedestrian phase per signal cycle. If you get unlucky, you will have to wait for both northbound/southbound traffic and eastbound/westbound traffic, which can take a minute and a half or more. But prior to changes made to ‘optimise’ city centre traffic flows, there were two pedestrian phases per signal cycle. The old signal phasing cut pedestrian delay by around one-quarter.
Reducing the cost of delay by 25% with one simple change: seems like a good opportunity.
Third, I was amused to see how these results were reported. For instance, here’s the headline in Stuff:
Auckland’s poor pedestrian flows cost $186m every year
Packing in pedestrians costing Auckland $186 million a year
These headlines are not accurate. As shown in the above figure, our estimate was that the annual cost of pedestrian delay on Queen St intersections was around $12 million. The discounted present value of delays over a 40 year period was estimated to be around $186 million. (This assumes no future growth in pedestrian volumes, so it’s likely to be a conservative figure for the present value of delays.) Furthermore, these delays aren’t necessarily an economic cost. Delays will affect some people who are on the clock (walking to meetings, delivering parcels), and some people who are out shopping or walking to work.
I guess this goes to illustrate how “cost of congestion” studies are reported in general – there’s a tendency to seize upon the largest number, ignoring relevant caveats about that figure!
What do you think about pedestrian delay in Auckland?
Processing...
Lols on media shoddiness. I know this is somewhat technical data driven stuff but surely journalists can spend two minutes to read through and clarify?
Also not quite correct that “public transport has catered for all of the growth [in pedestrians]”. Public transport has catered for a huge amount of growth, but of course we can’t forget the big growth in the number of people living in the city centre too. Them folks walk a lot.
That same research showed that there are now more people already living in the city centre than commute in by car.
Sorry, should have clarified: PT has catered for all of the net growth in people entering the city centre during the peak. Obviously there are a range of sources of growth in people spending time in the city centre.
Something that personally bugs me is when there are crossing points that could have the green man on automatically for extended periods without causing any delay to traffic (because of how the traffic signals / no turns are configured) but instead the designers persist with the buttons and the standard 2 second cross time per cycle. This forces delays on people walking who must wait at least one phase or often more (as the lights make you wait a whole cycle even if you push the button in the amber of the phase before). An example of this is crossing Wakefield St on Symonds St.
Yet curiously, the cross phase activates on demand in Newmarket (Mortimer Pass). Perhaps the council (or AT, whoever is responsible) should investigate where your change could be made.
I guess that the traffic engineers would want to model how this would modify pedestrian flows. That is, if they were instructed to expand their focus to cover peds…
When I was at Uni the police turned up one day to try and stop people walking across the end of Alfred Street at Symonds and promptly gave up as 100s of students kept crossing. About a week later AT finally changed the phasing to keep the pedestrian light green in every phase except for the *one* phase where pedestrians can’t actually cross and suddenly the intersection actually worked.
There are dozens of similar intersections around Auckland, particularly in the CBD.
Pretty idiotic from the Police when Alfred Street is a pedestrian mall so peds always have priority on it. Doesn’t work like that though, with no enforcement it’s full of parked cars and rat runners.
The pedestrian mall ends just before the lights, I understand why they were there as no one was getting across. Thankfully someone correctly identified that the signal design was ruining any respect for the law.
+1 I don’t know if anything’s changed officially wrt Alfred St but since the demolition of the Maidment Theatre started (and that of course has long since been completed), it’s been choked with parked vehicles all day every day. Presumably with the acquiescence of the University…
re Maidment – I just went past there the other day and was amazed to find it had gone. Why? Haven’t heard anything about a need for it to go, so was fairly staggered to find a pile of rubble there instead. What was the reason, do you know?
Yes: it was deemed an earthquake risk and not worth fixing. How exactly it was deemed an unacceptable seismic risk, while the rest of the Student Union – joined to it and of, to outward appearances at least, exactly the same construction – is still perfectly safe, beats me.
I swear the maidment burnt down too, so the seismic strengthening is required as well as the repairs, right?
The fire was a fair bit prior to the seismic assessment, I think, and apparently unconnected. The announcement of the closure and demolition of the theatre came very much out of the blue, from what I know.
Seeing as the Maidment Theatre was built well after the Student Union (10-20 years later ?), and was built to a far better standard, it seems like a load of bullshit that it was not up to strength while the Student Union was OK. It would be interesting to know the real reason.
I guess it is their building, so they can do what they like – but it was a great wee theatre, and I sincerely hope it is planned to be rebuilt, slightly bigger and more seismically betterer…. actually, was it theirs? Wasn’t there a significant amount of public money put into the building?
Wellesley/Kitchener on the side with the art gallery is a good example of this
I think AT should have a policy to remove all beg buttons from the city centre. How often is there actually a phase where no one is crossing? Probably only at night.
Fascinating and valuable research, thanks for sharing (and doing)! Really hope it leads to some action.
The CBD actually feels ok on foot compared to the suburbs, although of course it should be the focus as there are more people on foot. One of the problems I have with the New Network around East Tamaki is that transferring from the southern to eastern routes around Ti Rakau Drive either means taking your life in your hands on wide, uncontrolled, intersections or missing your bus while you wait forever at the lights to cross. Or you could spend an extra 15mins on the bus going to Botany and back.
The worst pedestrian delay I can think of is out east too – the crossing from Botany Town Centre to Kmart.
Really valuable point Martin – signals around PT interchanges or major PT stops should take walking movements into account. Especially when you’ve got to walk across the road to make a bus transfer.
I’m not sure if this is common practice or not – I suspect not.
BTC Kmart has to be the worst I’ve seen in Auckland.
I’m not out that way very often, though when I am I’ve usually got the dog with me. There’s no way I’ll walk the dog across that intersection, it just doesn’t feel safe enough. Particularly with the slip-lanes.
+1 missed transfer on GNR at Pt Chev the other day due to the really slow ped phases.
Yes that one bisects the town centre too, which is one of the reasons customers go elsewhere. It must be having a big financial impact on retailers. Also means there are loads of people not waiting for the green phrase; lots of dodgy crossing being modelled to children.
I thought running the Barnesdance twice each cycle was a huge improvement. It didn’t really impact on cars that much but reduced the waiting for pedestrians heaps. Not sure why the gave it up.
Sorry, but I have to ask: what is “active transport users”? It’s presumably not people in cars or buses, so, then: what is it?
It’s generally used as a shorthand for people on foot or bikes. Agree that it can be a bit obscure!
People on bikes, people walking, people on skateboards, people in wheelchairs etc.
And equestrian is often counted, but not particularly relevant here.
“f**k your Honda Civic, I’ve a horse outside!”
ummm, thanks Daphne, i think. Not entirely sure why they have plastic bags on their faces, but then again, there’s a lot about the modern world that I really don’t understand these days. Like, why do people watch the Block? Why do Americans vote for Trump? Why do people wear onesies? And why do girls have rips in their jeans? Sigh. So many questions.
Peter, Simon, Sailor – thanks. I’m evidently an active transport user myself. Didn’t know that till just now (except for the horses). But i still think it would be easier to just have a legend that said Pedestrians, Cyclists etc, instead of Active Transport users.
Cheers for the video Daphne, been listening to a few of their tracks through the day as a result.
Guy, ‘active user’ is a really commonly used catch all term, I’m surprised that you haven’t heard it before!
The song was released about the same time Ireland had a problem with up to 20,000 “trophy” horses released into the wild and having to fend for themselves. One of the more weird effects of the Celtic Tiger and its slaying.
https://www.theguardian.com/world/2010/oct/15/ireland-homeless-horses-cull
When you’re talking about light phasing, you noted: “On Queen St, there are around four times as many pedestrians as vehicles. On High St, there are around 13 times as many pedestrians as vehicles”
This is funny really and quite informative. The old adage is that “what you feed, grows”. If pedestrian flows have grown this fast, without being fed (ignoring residential and commercial pressure for now) – How much can we grow the ped count if we start feeding it?
It’s important to me that we make the city more pedestrian friendly, if not for the lower carbon cost and intrinsic health benefits, then at least for the commercial benefit that comes from more people walking by your shop without being concerned with traffic.
Tangential to that last point, anybody else frustrated when folks say that malls do well because of free parking? I’d have thought that malls do well because you don’t have to cross roads or watch traffic once you’re there.
Absolutely. I can’t stand them but lots of people quite tightly point out that they can be much more relaxed with children when they’re not having to restrain them all the time.
Ah come on Heidi, children are wonderful.
🙂 Ha. I thought I might get a comment about my “tightly” typo. Didn’t spot that grammatical ambiguity. Children are indeed wonderful. Even other people’s.
Especially other people’s
Malls do well because the “public space” is actually private space, and so people feel safe there. My favourite quote on this, from Elizabeth Farrelly in “Blubberland”, where she notes:
“The contemporary shopping mall is a classic instance, striving to ape the complexity and interest of a traditional market, but with one, overarching difference; the space is privately owned and entirely under private control. The air is cooled and conditioned and gently muzak-laced; the lighting, level changes, transparency and detail are minutely orchestrated to slow your gait within microsends and lull you into buy-mode; and no one, other than the owner, has any right to be there. Everyone else is there on sufferance. Anyone undesirable, or homeless, or obviously poor, can be summarily shuffled out so that the mass-narcosis ofconsumerism never pops.”
“it is interesting to note the increasing interiorisation of shopping as an activity. From the traditional street market where interiors were ususally makeshift and temporary, to the modern mall where entire city blocks and precincts are interiorised to give the (usually female) shopper the illusion of being in a vast, sparkling, bejewelled, cathedral-like home. Woman like malls because they’re known, comfortable, clean and safe – from muggers and spitters, from sun, storms and mendicants. Here, at fantasy-home, women will relax, and when they relax, they will spend.”
“Malls titillate and relax, even while they make you feel needy, inadequate and dreamily disoriented. They’re meant to. It’s like chocolate. If you can be made to feel bad, in a small way, the more want soothing, and the more you buy. One thing you don’t see much of in a mega-wall, therefore, is social life. Whereas in a high street you might stop for a coffee, in a mall you bump into someone, you say hi and press on. This is because, from the first car-park moment, the place is designed as a disconnect, separating you from your reality and from your higher, warmer self. It’s designed to put you in a bubble – a car-like bubble – of self-gratification.”
Did I inherit a different gene? My pleasures in malls are limited to:
Walking out with exactly the items I went in for, and no more.
Going by bus in the week before Christmas and smiling at every stressed face.
Ok agree with opportunity for mall up and down length of Queen street BUT a light rail line as well? Doesn’t the CFN2 say this is not a tram but a rapid transport light rail?
A nice slowish pedestrian friendly tram, hop on/off would be good but a commuter RTN rail? How will this be conducive to a pedestrian friendly safe mall?
Having a slower speed on the light rail line down queen street would answer your question. 3 stops down the length of lower queen street would be adequate I’d guess.
“How will this be conducive to a pedestrian friendly safe mall?”
24 vehicles per hour, travelling at 30 km/h: far fewer than any of the shared spaces.
The same vehicle could do both – operate as an RTN in outer areas and as a tram in Queen Street. These days most LRVs on the market offer a low floor design that is good for pedestrian access. So you could have an LRV operating at a higher speed (80 to 100 km/hr) in outer suburban areas that would slow down to, say, 30 km/hr in Queen Street in a kerbed environment with widened footpaths or 10 km/hr in a pedestrianised zone. This has been recognised as one of the principle advantages of city centre LRT – you can clear out the traffic and open the street up to pedestrians far more. See this conference presentation:
http://onlinepubs.trb.org/onlinepubs/conferences/2015/LRT/DBertrand.pdf
In addition to the other comments, I wonder what the impact would be of the rail line being encompassed with either a low curb (with ramps where boarding is allowed) or a drastically different texture, to alert the “distracted” among us.
It has to be better than the diesel buses they operate as noisy ‘hunter-killers’ on Queen Street.
Almost _anything_ would be better than that! 🙂
Peter estimating pedestrian delay versus the free flow alternative is relatively easy to research, so I can see why that is done. But what about looking at access? Isn’t access the really important issue i.e. how many people can get to their desired destinations in as short a time as possible. The Bertaud paper I quoted in my yesterdays article indicates there is big productivity gains to be made by increasing the number of people who can access their desired city locations as quickly as possible.
Of course modelling access would be fiendishly difficult -wouldn’t it?
But surely access (not speed per se) would be one of the main benefits of light rail/pedestrian Queen St? It would mean the 40,000 people who arrive in the CBD by public transport, or who live in the CBD are able to access a bigger proportion of the CBD in as short a time as possible. Further that increased access in as short time as possible would induce more travellers to use this travel option and access those locations.
Modelling accessibility isn’t that hard. (There are some technical wrinkles to modelling pedestrian accessibility that make it a bit more complicated.) But the first thing that you realise is that, holding land use constant, places get more accessible when you can get there faster. So delays matter for accessibility.
Some folks from Auckland Council analysed accessibility to employment by walking within the Auckland city centre. We reported that research here: https://www.greaterauckland.org.nz/2017/07/27/research-pedestrian-connectivity-economic-productivity-aucklands-city-centre/
Interestingly, there was a quite strong positive association between walking accessibility and labour productivity.
In some separate, unpublished work for an Australian city, I modelled changes in pedestrian accessibility as a result of putting new pedestrian links in place around the CBD. Even if you adopt conservative assumptions about agglomeration elasticities, the resulting productivity gains would be equal to a substantial portion of the cost of the projects – around 50%. Direct benefits to users from more convenient journeys would be additional to that.
Of course, we *would* expect land uses to change as a result of changing walkability. Stu Donovan’s got a post about that for tomorrow…
Ooh goodie. How much of this sort of research is done for suburban centres?
Not a lot to be honest. I think more should be done, though.
I tend to see the city centre as the great ‘demonstration project’ for suburban centres – it gives us a chance to see what works in terms of PT access, walkability, public realm investment, etc.
It’s dismissed by many (eg suburban retailers) as irrelevant, though. So, yes, I’d love to see some suburban work done. Even if it’s just an exercise in determining the relevance or otherwise of cbd work.
Great stuff Peter and looking forward to Stu’s article : )
This research measures connectivity (effective job density).
https://www.greaterauckland.org.nz/2017/07/27/research-pedestrian-connectivity-economic-productivity-aucklands-city-centre/
Thanks Peter, it’s great to measure the costs of pedestrian delays. I think giving pedestrians higher priority at inner city interesctions is something engineers have long suspected makes sense, but without the hard numbers it’s been difficult to justify it. Hopefully your research can begin to change that. “what gets measured gets done” and all that…
Peter,
Interesting that for the Queen St intersection surveyed over 50% of the pedestrians are crossing Queen street on to the other side of the road. Either directly or diagonally. Which is a lot of crossing activity.
Any idea as to the likely delays pedestrians will face when crossing Queen St either mid block or at intersections, when we have Light Rail in place?
Seems to me that LR will in its own way present nearly as big a block to crossing Queen St than happens now with the present layout. Given the length of the LR vehicles being proposed and frequency you are talking about a significant moving “wall” presented by a LR Vehicle” whenever it passes, more akin to a stream of buses than a dinky Tram of old. And with them running both ways, could end up with parts of the road can’t be crossed as easily as the renders would suggest.
So won’t pedestrians potentially face a double whammy of [shorter] intersection crossing delays like they do now, and also delays whenever a LRV [or two!] is crossing in front of them, when they want to cross the road anywhere else.
Of course, if pedestrians can cross any time there are no LRVs coming then that reduces the demand at intersections. But if that mid block pedestrian traffic also forces the LRVs to have to significantly slow down along Queen St, then that means each passing LRV becomes a bigger and slower moving block in terms of the time it takes – further delaying any crossing of the road – which as we see is 50% of the pedestrians on Queen St currently.
So maybe we’ll end up with a fence down the middle of the road to prevent folks doing mid block crossings.
12 vehicles per hour per direction travelling at 30 km/h block the road for 96 seconds, or 2.6% of each hour. In the worst case there will be two straight minutes with no vehicles and then 12 seconds with one vehicle. In the best case, there will be over 4.5 minutes between vehicles.
There are currently 18,000 vehicles a day, which would equate to approximately 900 per direction in the busiest hours. Delays will be orders of magnitude less than what we currently experience. The biggest issue is that there will be so few vehicles that pedestrians may forget!
Patrick had some observations on a similar transit mall in Melbourne: https://www.greaterauckland.org.nz/2017/07/10/democratic-tin-observations-bourke-st/
The pictures say it all, really – very close to free-flow conditions!
With the LRT in place you would not need a barnsdance phase. You could go to a simple two phase system where either the LRT flowing (with north south peds) or Victoria Street (with east west peds). This could be on a short cycle, say 60 seconds, so pedestrian delay should be greatly reduced.
There are few cities I know where Light Rail corridors in urban centre areas are fenced. There ARE indeed some where the “wall of trams” in the town square etc is a real problem – but we are at a guess, 10-20 years away from getting a high enough frequency for that to happen, I would say (a single Dominion Road line will not cause that issue).
Removing vehicle traffic at all other times except when trams go is a pretty major bonus, which will only be partly eaten up bu the tram blockages.
Nice research Peter! The obvious corollary is that there must be some CBD streets where pedestrian crossing times are slow so that the level of pedestrian activity may be suppressed as people take alternative routes. I was thinking that Fanshawe Street and parts of Beach Road might be examples of this. The pedestrian crossing times at signals on both are painfully slow.
Thank you Peter.
Do you think that there is any case for protecting those pedestrians waiting at the lights from the inclement weather we have been experiencing?
Some of them are much better than others.
This consideration also needs to be given priority with the final plans for the CBD bus interchange.
Yes, good point. At times in heavy downpours, I pick my way through the cbd based on where there is good shelter. It’s pretty limited.
I support your idea Peter such as helping pedestrians and to stop catering to cars. But I disagree with some of the details.
Some Queen St intersections ran two Barnes Dance phases per cycle. This is true. What is also true is that they ran a 180sec cycle. Now they run a 90sec cycle with one Barnes Dance phase. On average, there is no change in pedestrian delay. In fact, due to some technical quirks, I expect the average delay has actually decreased slightly. I don’t blame you for reporting on what you have been told, but what you have been told is not completely correct.
You say pedestrian delay costs 12m a year. This is fair enough.
But what of the cost of delay to bus passengers stuck in buses at these same intersections? There are major bus routes through many of these intersections. This is a zero sum game. There is finite time. If you give it to pedestrians then you have to take it away from buses. You cannot says “Look at all the delays to pedestrians!!!” without also looking at the other side of the equation, delay to people on buses.
I think on balance there are more pedestrians so the BCR would still be positive, but it is not good to show an incomplete picture.
If you really wanted to reduce pedestrian delays, get rid of the Barnes Dance phases. Your surveys suggest the majority of people don’t want to cross diagonally, so the Barnes Dance phases is not needed. Those things are very time inefficient. Ban some turns and just run a north/south only phase and a east/west only phase and you could run a two phase 60sec cycle intersection. Pedestrian delay would be dramatically reduced as well as bus delay. Win/win for everyone except the drivers who want to turn. I expect this is exactly what will happen when we run trams down Queen st.
The length of a full phase was reduced from 120 seconds to 90 seconds when they cut the second Barnes dance out. We took that into account in our analysis.
We found that 23% of pedestrians crossed the Vic/Queen intersection diagonally. Removing the Barnes Dance would immediately delay them all by 30 seconds or so, as they would have to wait for the lights to change.
Keep in mind that there are around four times as many pedestrians as vehicles on Queen St. So delaying 23% of pedestrians by ~30 seconds would be as costly as delaying *all* of the vehicles by ~30 seconds.
TLDR: Barnes dances are great.
They are, and double-phased ones are awesome.
The maximum cycle length was 180s, not 120s. I’d like to see proof that removing the barnes dance would increase their delay. With the right phasing, there would be almost no difference in average delay because the total cycle length would be reduced without the Barnes Dance phase.
“Great” is just an opinion. If you are not crossing diagonally, they are totally unnecessary and increase your average waiting time if you didn’t have them. This is a fact. No opinion on “great” can negate that. If 1/4 of peds get an increased delay at the price of the other 3/4 of peds getting reduced delay, you still get a net benefit and my point still stands. If you have low diagonal crossing demand, a Barnes Dance is a political feelgood exercise that actually costs people time. Having said that there are also safety benefits, but that is an entirely different point.
My point is not about vehicle numbers, but bus numbers. Queen/Wellesley is a great example. Major bus routes through there. If you have 1 bus with 50 people stuck on it waiting to get home, it adds up. Did you count the delay of all the people in buses?
I go back to my other point. It is a zero sum game. If you give time to pedestrians, you have to take it off bus users. Still a benefit for train/ferry users and those who live in the city.
Given that I see so many people on their phones while waiting at the lights, on the train, on the bus, in the car, at home on the couch, in bed while watching tv, is the delay actually a real economic cost at all? They are filling that time with what they deem is of value to them.
If you save time at one intersection, you may end up losing it all waiting at the next intersection. Overall all I think these numbers are incomplete and a bit silly, but I totally agree we need to focus more on improving things for pedestrians and the first thing is to measure the current situation and give a complete picture. For too long we have only measured cars and thats what we have prioritised.
It’s a zero sum game *only if* there are equivalent numbers of pedestrians and people in vehicles. On Queen St, and a variety of other places around the city centre, that’s not true. In these locations, giving more signal time to pedestrians benefits more people than giving the time to cars.
With regards to the critiques of incompleteness: You’d obviously have to consider overall delays for all users when making signal timing decisions. However, we didn’t analyse delays for vehicles, because the point of the research was to measure something that hadn’t been measured before. Delays for vehicles are already commonly measured and modelled, so we didn’t see the point in duplicating existing work.
Finally, your assertion that Barnes Dances increase delay relative to other signal phasing strategies isn’t borne out by the data. You can see that pretty clearly in the comparison between Queen/Vic and Queen/K Rd, which doesn’t have a Barnes Dance. Pedestrian delays were higher for Queen/K Rd.
Drilling into the details, the current Queen/K Rd phasing is no faster for people going straight through the intersection, and 27 seconds slower for people going diagonally across the intersection. In short, Barnes Dances work better for people walking.
Which is what it feels like on the ground. Having to wait for two phases feels slow.
A few Sundays ago we were eating Popa’s Pretzels on the corner of Queen St and Wakefield St intersection. (Their vegan manakish is fantastic.) The intersection seemed to have two BarnesDance crossings per cycle. I had a couple of 11 year olds with me and we took turns counting the number of people in the cars and the pedestrians at each phase. Even that far up Queen St, there were about 4 or 5 times as many pedestrians as there were people going through in cars.
The uninitiated 11 year olds thought the cars were getting too much time.
The famous Shibuya intersection barn dance is great 🙂
Thank you for an interesting post with the significant sentence being “But there’s also an opportunity.”.
These are useful statistics you are reporting. Is it worth while attempting to distinguish between tourists and non-tourists with the latter walking purposefully from place to place while the former tend to be elements of white noise randomly delaying the purposeful pedestrians? Queen st might be just about acceptable for footpaths if there were no tourists; since we have tourists and they seem to be increasing then it is obvious that most of the traffic should be removed from the CBD.
The Victoria st and Wellesley st crossings are painful so I prefer jay walking across Queen st between junctions; this is fairly easy because the traffic is so slow and congested. Is any attempt made to count similar jay walkers or are we insignificant?
Peter, the article raises a really important issue. Almost all of our transport models do not include pedestrians. We inherently bias ourselves to the determination of traffic effects.
Can you also measure the cost of delay to the vehicle using those intersections?
With a comparisons we can find the priority.
Yes excellent work Peter.
I am surprised that there are roughly the same number of cars entering the city as there have been for a while. To me it appears that there are less. Perceptions are not always accurate.
While the focus of your article is on a better outcome for pedestrians, with which I agree, the figures also strongly suggest that there should be a much better outcome for public transport users. While a person in a car can avoid traffic by taking an alternate route that same option is not open to bus drivers. AT needs with some urgency to address the issue of bus lanes. Have they caught up their missed target of last year? How are they progressing against this years target?
Generally AT seem to be ineffectual at many things that they do. Twice in two weeks there have been pieces on Seven Sharp about the shambolic state of one of their car parks. Unauthorised parking around transport stations is absurd and continues. For example,on occasion parked cars at Akoranga prevent bus access. Aucklanders are saying they want, and they deserve, a far better public transport outcome than they are currently receiving.
Clearly the best solution would be to start charging people for standing space. Currently you can stand at a bus stop or at a pedestrian crossing for free. That results in too many people wanting to stand at the bus stop or street corner and too few standing spaces. Standing is rivalrous because if you stand somewhere that means someone else can’t. It is also excludable because the technology already exists to have physical barriers like turnstiles or we could use the GPS function on smart devices and directly charge a fee every time someone stops on a footpath. Being both rivalrous and excludable means standing space is actually a private good. The efficient way to allocate private goods is to price them at their marginal cost.
+1, then we can charge them at the same price per square metre as the vehicles and actually achieve the economically efficient distribution of CBD space 😉
Miffy maybe we should achieve the ‘big gains’ from spatial economics (such as correctly pricing roads and parking) before we take the concept to the absurd level of pricing the much smaller amounts of land used for footpaths, bus stops etc?
This is a great piece of research. Does anyone know – anything underway in “NZ’s Most Walkable City”?