An article in the New York Times looks at the recent announcements that transit (PT) ridership in the USA in 2013 was the highest since the 1950s – much the same as in Auckland. What’s perhaps most interesting in the 2013 numbers is that petrol prices doesn’t seem to have featured as much in the reasoning behind the increase:

The trade group said in its annual report that 10.65 billion passenger trips were taken on transit systems during the year, surpassing the post-1950s peak of 10.59 billion in 2008, when gas prices rose to $4 to $5 a gallon.

The ridership in 2013, when gas prices were lower than in 2008, undermines the conventional wisdom that transit use rises when those prices exceed a certain threshold, and suggests that other forces are bolstering enthusiasm for public transportation, said Michael Melaniphy, the president of the association.

“Now gas is averaging well under $4 a gallon, the economy is coming back and people are riding transit in record numbers,” Mr. Melaniphy said in an interview. “We’re seeing a fundamental shift in how people are moving about their communities.”

New Zealand has been different in regards to petrol with prices hovering around or even above the peaks of 2008 for the last few years. Here’s the graph of our average weekly petrol price up to 7th March.

Petrol Price - Mar 14

Furthermore, the article notes how over the last 18 years, PT ridership has grown faster than the rate of population growth, whereas the level of driving per capita has fallen:

From 1995 to 2013, transit ridership rose 37 percent, well ahead of a 20 percent growth in population and a 23 percent increase in vehicle miles traveled, according to the association’s data.

Stronger economic growth is playing an important role in the increased use of public transit, as more people are using the systems to get to an increasing number of jobs, the association reported, and transit agencies are nurturing growth by expanding their systems or improving services.

“We’re seeing that where cities have invested in transit, their unemployment rates have dropped, and employment is going up because people can get there,” Mr. Melaniphy said.

Overall public transit ridership increased by 1.1 percent from 2012, with the biggest gains in rail service and in bus service for smaller cities.

Here’s what Auckland’s vehicle kilometres travelled per capita looks like compared to the number of trips per capita. Note: we won’t have VKT data for 2013 till later this year, also the latest PT trips per capita has started climbing again.

VKT vs PT Trips per Captia 1

There’s often debate about whether the levelling off of traffic growth and the fairly dramatic increase in PT use over the past few years is a “blip” – caused by the global financial crisis and the fairly long recession that followed it, plus highly fluctuating oil prices in the past few years – or whether the changes are a longer term trend. With US ridership booming despite lower oil prices and at a time of growing economic success, the “blip” argument seems to be getting weaker and weaker. The longer term trends are well summarised in the article’s final two paragraphs:

Todd Litman, an analyst at the Victoria Transport Policy Institute in Victoria, British Columbia, said the new data were the latest indication of changing consumer preferences as a result of increasing urbanization, an aging population, and environmental and health concerns.

“A lot of people would prefer to drive less and rely more on walking, cycling and public transit, provided that those are high-quality options,” Mr. Litman said.

Time to change those traffic projections NZTA.

Share this

31 comments

  1. The trend, always the trend. That’s what matters in forecasting. Status Quo bias, assuming current conditions, or worse past ones, are permanent, needs to be constantly tested.

    1. The most generous conclusion is ‘cognitive dissonance’. The inability to see facts because they contradict deeply held beliefs.

      Ha, just thought I should look it up and found this:

      “When dissonance is present, in addition to trying to reduce it, the person will actively avoid situations and information which would likely increase the dissonance” [1

      *Scenes of the MoT and Ministers grimacingly averting their eyes from traffic and ridership counts*

  2. Its all part of our clever plan. We all know in our hearts that things are not what they were, this time will be different, history is bunk, we are at a ‘tipping point’, its a generational change, but we rely on those regular cheques they send us from Detroit so we have to say things like public transport is an inferior good, given a choice people will opt for cars, 1/2 of the future trips to intensified areas will still be by car, if there is no parking maybe people just wont choose to build we dont really mean any of it.

  3. Ok so Patrick’s comment on dissonance made me exam my views on this in a bit more detail. The second graph is very pretty but presents some assumptions that haven’t been made clear. First the lines use different scales but appear to assume that it is fair to compare 2 PT boardings with vehicle 100km travelled. That implies a trip length there of 50km and back of 50km which by any standards of sprawl seems huge. Second neither line has the origin or zero in the scale so that distorts the ‘trend’ . Third I am not sure who collects data on vkt per capita but I assume it is based on the http://www.transport.govt.nz/ourwork/tmif/transport-volume/tv001/ data. The data that notes for Regional vkt “Quality of the data – The counts and estimates in RAMM are updated periodically so the estimates may not reflect either traffic growth, or responses to events such as oil price shocks”

    Thats right may not reflect traffic growth! Ramms is a database used for road maintenance. It includes the length of road segment and an estimate of traffic flow. Multiply and add and you get a very rough estimate of vkt for the region. The problem is the estimates are not counts for most roads and are not reviewed each year. For local roads they will often be an estimate such as 100 or 500 and only changed if the character of the road changes noticeably. Arterials are counted but not each one every year. When a recession hits the arterials go down a bit. If you take these fairly static numbers and divide by an estimate of population you will get a line something like the blue one. I dont think it is dissonance to look at that and not believe that PT boardings have grown by four times what vkt has dropped. Yet that appears to be the message of this graph.

    1. We have very good data for VKT nationally, based on your odometer value which gets reported whenever you get a WoF/CoF. The notes suggest that the regional estimates don’t add up to the national total, which seems bizarre. Why wouldn’t you scale the rough regional data to fit the accurate national data?

    2. But, but, mfwic, PT trips per capita have risen 30% in a decade. 30%! And VKT is down a full 2%. Forget PTKT, that’s irrelevant. It’s game over for cars really.

      But seriously, a graph like that needs a lot more explanation and a modicum of logic. Usually that type of scaling is used to identify trends such as the number (not value) of house sales vs interest rates, or mortgagee sales vs interest rates, where there’s a reasonable expectation of a measurable inverse relationship because they’re so closely linked.

    3. The lines use different scales so that the initial values are the same in order to show the relative trend. Anyone with basic knowledge in statistics would understand what the graph is trying to show.

      1. Hi Sailor Boy, thanks for the ad hom (I thought those were deprecated on this blog). Did you even read my second para?

        1. I read it but decided not to address it.

          Also this isn’t an sd hom. Knowledge of statistical practice is pretty rare tbh. A lot of people either wouldn’t take statistics in year 13 or remember it if they had. So to imply that you wouldn’t understand due to a lack of education is hardly attack.

        2. Apology accepted SB. I agree that most people have a limited grasp of statistics (such as puzzling over the meanings of margin of error or standard deviation). While not a statistician, and currently an infrequent user of stats, I did study stats to a fairly high tertiary level (initially in engineering, but in greater detail in financial management). I’m not for a moment suggesting that the presentation was intended to deceive; the problem is the visual impression it gives to non-statistical people.

      2. Actually, I see no logic at all with the scales. If you want to show relative movements, then the scales need to both be on the same relative scale, which in this case they’re not (the PT scale is way larger relatively than the VKT scale – if they were on the same relative scales the VKT would be close to constant).

        If you start both scales at zero, then it’s pretty clear that PT trips are up a whole heap, while vehicle kms is pretty much constant. That would give a more informative graph with respect to the actual movements.

        1. Actually this is my one concern too. The scale of vkt should really go from 6k to 10k so that the end scales are matched at both ends, but this is antithetical of the change suggested by previous commenters

    1. That is a cool looking graph, but they forgot to multiply the bus and train numbers by two to get boardings then multiply by 365 to get annual boardings, then put PT boardings on a different y-scale, then crop the graph and alter the y-scales until the bars for rail and bus are higher than everything else.

      1. > multiply the bus and train numbers by two to get boardings then multiply by 365 to get annual boardings

        Both of the statistics in this post’s graph are annual: annual VKT, annual boardings.

        > then put PT boardings on a different y-scale

        VKT and boardings are measured in different units (i.e. kilometres per person, and a plain number of events per person). You can’t put them on the same scale.

        1. Just ignore me Michael I was being a smart ass setting out how someone would doctor the data to arrive at the conclusion that cars are not actually important despite evidence they are.

        2. mfwic, as I’ve discovered to my despair humour doesn’t go down too well on this blog. But you’ll be pleased to know that I got it! I was going to be a wiseguy too and point out it’s about getting TO work only. No one knows how they get home. But I won’t.

        3. Thanks jonno1 next time I will deconstruct it up front and just type “A man went into a bar, the barman said something funny – and everybody laughed”

  4. As long as both graphs are showing on a per capita basis, I don’t see the issue. Even if the two things being compared cannot be compared directly (VKT and PT Trips), you can still compare the per capita basis. The only scale that matters is the time scale. As long as that is constant for both then you can can compare the growth in PT vs the decline in VKT. The 0 starting point is also irrelevant.

    As for micro cars, its a non-issue. The market will determine the outcome and my bet is that no-one will be interested in buying a micro-car for $80k. What a waste of money.

  5. Widening roads when there’s a solution available to build and lease narrow, quick, and nimble single occupant vehicles comes at an incredible cost to the citizenry. A cost analysis will find that widening roads is far, far more expensive and therefore, the most horrible waste of money.

  6. Except no-one here is advocating road-widening because of the obvious cost reasons you point to. Improving existing PT services is a far greater use of public funds as opposed to creating some subsidised fleet of private vehicles that are only useful for single people with long commutes. You would need a huge uptake of those vehicles in order to get even marginal benefit to the whole network.

    1. You can also buy a narrow, quick and nimble single-occupant vehicle right now, for a couple of thousand, not $80k. They’re called “motorbikes” and they’re even less popular than pushbikes (for Census-day trips to work, at least).

      1. Sure, motorcycles have been around a long time, but unfortunately that haven’t caught the imagination of the vast majority of commuters. A right-sized narrow four-wheeled car capable of lane-splitting with standard car seats and doors, a roll cage, steel bars on the doors, heat, air conditioning, and (obviously) weather protection is not the same as a motorbike.

        A new commuting tool exists. How do people make use of the tool? Given people’s obvious interest in single-occupant driving, perhaps offering it as a pt option in order to save open space and the costs of road widening is the right method.

        1. “Single occupant commuting is still far and away the most popular form of home/work/home transport in the US…”

          “…perhaps offering it as a pt option in order to save open space and the costs of road widening is the right method”

          For me, these two quotes summarize my hesitation about the various narrow car proposals that come up from time to time. If we assume that single occupant commuting is in fact the preferred form of transport (a claim that is at least true for some subset of the population), then it seems to me that offering it as a PT option would remove the very thing that makes it attractive in the first place, namely that its no longer single occupancy.

          I’m not one of the people who prefer single occupancy commuting, so feel free to factor that is as bias in my opinion, but I still fail to see how narrow-vehicles will change anything. Even if we neglect the much improved ability of PT to move volumes of people at peak times, and space considerations at various destinations, it still seems to me that this kind of scheme fails in delivering the single occupancy experience that is apparently popular in the first place. This is before we note that until some critical percentage of the vehicle fleet consists of these cars, the total effect will be near zero.

        2. Like bike sharing, narrow car drivers could rent/lease narrow cars from the government or buy one outright, so I think the pt part of the equation wouldn’t scare folks off.

          I think the added value attributes of narrow cars – maintaining single occupancy, faster commutes, easier parking, low electric power costs, quick, nimble, fun-to-drive… – would attract lots of interest from commuters.

      2. I’d love to ride a motorbike but getting a license is prohibitively difficult and so are ACC charges and registration. Punished for other people’s habit of killing motorcyclists

Leave a Reply

Your email address will not be published. Required fields are marked *