Further patronage analysis

I must admit I got a little sidetracked in writing yesterday’s post about PT success, by the analysis I did of Auckland’s bus patronage since 2002. Analysing the patronage of the bus network, and of all PT modes as a whole, is actually something deserving of its own dedicated blog post as there are some quite interesting things you find if you look closely enough. So I’ve updated and extended the Excel Worksheet to include all patronage, and I’ve also made some distinctions between patronage figures just above/below average and those well above/below average.

If we start with the bus patronage again, I have updated the table I put together yesterday to include estimates for October, November and December 2010. Hopefully over time Auckland Transport will release the patronage information for these three months – and I will be able to update these with the real information. But I think I’ve taken a reasonably good approach to my estimates – working out the increase in general for 2010 against 2009 for January to September (5%), then for November and December 2010 applying that increase on the 2009 numbers. For October I had to do something a little different, as back in October 2009 there was the bus lockout – so I’ve applied the 5% increase to the October 2008 stats, potentially an under-estimate but I can live with that. This has also enabled an update to the per capita statistics to include 2010 – as full yearly patronage (though only estimated for 2010) can now be included. That brings us back above 34 bus trips per capita per year for 2010 – which is a big improvement on recent years but still below 2002-2004 figures. Some modeshift to train appears the likely reason behind this.Matt L pointed out in a comment on yesterday’s post that some of the declining bus use around 2005 might be attributable to that being about the time when we got the first “SA Trains”, and about the time that rail patronage really took off. While that’s probably true to some extent – looking at the total patronage statistics for 2002-2010 shows that 2005 to 2007 were still pretty bad years. Total patronage fell from 53.1 million in 2004 to 50.1 million in 2005, and it took until 2008 before patronage was once again higher than it had been in 2004.
On a much brighter note, it’s clear that total patronage has gone up massively in the last few years, meaning that the early 2002-2004 years now appear as being ‘below average’ years quite clearly. This year in particular, the patronage statistics have been particularly impressive. For October, November and December I have used a similar method to what I said above for bus statistics – though I used a 6.2% increase instead of a 5% increase: because PT patronage as a whole in the first nine months of 2010 was 6.2% above the same months in 2009.

Some interesting things jump out at me from looking at these numbers. Perhaps one of the stranger statistics is how high August patronage was in 2002-2004 compared to March patronage; but how that has turned around in the last few years which March becoming clearly the busiest month of the year. The timing of Easter may have something to do with March being so high (Easter has fallen in April for the last two years), but it still seems strange that August’s total PT stats back in 2003 were only matched in 2010. For bus patronage, 700,000 more bus trips were taken in August 2003 than were taken in August 2010. That’s really strange.

Looking at per capita PT boardings, thankfully the big patronage increases in 2010 has pushed us over 43 trips per person – an increase from 41.6 trips per person back in 2002. I’m debating whether I should use Statistics NZ’s population figure for the “Auckland Region” or simply that of the “Auckland Urban Area”. I have settled on the region as a whole, although the table below shows a comparison of the two in terms of PT trips per capita. It doesn’t really make too much difference in terms of the change over time.One aspect of the data that I am a bit wary of is that I have got all these figures off the line graph that ARTA included in all their monthly business reports: This means that there’s probably a little bit of impreciseness in the figures, as sometimes it’s a little tricky to tell whether I should say the patronage was 4.2 million or 4.3 million. This shouldn’t make too much of a difference to the longer term statistics, but it might mean that some months come out as being below average or just above average incorrectly. To compensate for this, and also to provide quite an interesting further insight into the data, I’ve gone through both tables (bus patronage and total patronage) and changed the colours of months where the patronage was at least 300,000 lower or higher than average. This highlights what might be called ‘particularly good’ or ‘particularly bad’ months.What comes out most obvious again is just how bad things were in 2005-2007, and also how much patronage has increased in 2009 and 2010. I really am curious to know what happened in late 2004 and early 2005 to cause such a dive – something that wasn’t really arrested until the high petrol prices of 2008 kicked in.

Another way (in addition to how a particular month compares with the average figure for that month over the nine years) to assess the ups and downs of the statistics is to compare how a particular month performed against that same month the previous year. Obviously, as I don’t have any statistics for 2001, I can’t compare 2002 months against anything earlier, but for later years the results are quite interesting indeed – especially when you apply the same “particularly good/bad” additional measure as I did in the graph above. The results are shown below: What becomes clear in the above table – more so that earlier tables – is the timing of the fall in patronage in late 2004 and early 2005, and then the timing of the recovery, firstly in 2008 and then again pushing forwards in 2010 yet again.

In fact, one of the most interesting things about all this analysis is what appears to be the immense impact of the May 2005 bus stike. During that strike action, Stagecoach buses were off the road for around six days – plus there were further strikes before that big one (and maybe afterwards, I can’t remember). Interestingly, by contrast the October 2009 bus lockout did not have anywhere near the adverse effect on patronage in the long term – perhaps because people switched to catching the train for a few days rather than going back to driving? While it seems PT patronage in general was declining before May 2005, the biggest falls (August 2005 bus patronage was a massive 1.2 million less than August 2003) appear to happen in the few months immediately following the bus strike in 2005.

Looking a bit more longer term, across the whole of the nine years the data tracks, there are some interesting statistics when you put it all together: Unfortunately it’s a bit difficult to separate out the rail and ferry data, so I’ve had to lump them both together. What immediately stands out is how Auckland’s public transport system has diversified over the last nine years – back in 2002 almost 88% of PT trips were on the bus, whereas in 2010 that figure had declined to just under 79%. As ferry patronage has been fairly stable, most of the change has come as a result of the massive increase in rail use. This is reflected in the 111% increase in ferry/train use between 2002 and 2010 (almost all of which appears to have been on the trains). This compares to only a 8.6% increase in bus patronage over the nine years.

Looking at things from a per capita basis, it’s heartening to see that there has been an increase in per capita PT use in Auckland since 2002, even if it’s only relatively small (4.2%). What continues to worry me is the 6.6% decline in per capita bus use over the period of the data – a reflection of the fact that, aside from the central connector and the Northern Busway, we’ve actually done very little to improve the bus system over the last 9 years. That’s where I think our real focus needs to be in the immediate future – particularly if we want to achieve ARTA’s long-term goal of 100 million trips across the whole network by 2016. A big chunk of those are going to have to be on the bus network – and they won’t happen by magic.

13 comments to Further patronage analysis

  • Chris

    Great commentary and analysis! You aren’t by any chance moonlighting as AKTransport’s resident business analyst?

    It is interesting to see the general trends, and I agree with your conclusion – we need more of a sharper focus on bus transport.

    I wonder, what has the contribution been of the 40% discount scheme for tertiary students? Would that account for any significant numbers?

    • I agree there needs to be more focus on bus travel — however I disagree with focusing on bus travel for the sake of increasing bus patronage. A better question is where can we best use buses?

      For example, I live in Mt Albert and work in Avondale. Some days I bus or drive to work and thus get to observe the New North Road bus service. I note that most (if not all!) bus service operating between Avondale and Mt Albert during the morning peak carries less than 10 passengers each.

      I recently managed to get hold of the New North Road bus timetable and I was shocked to see 19 buses terminating into the city between 0800-0859. That’s a bus every 3 minutes for a route which probably only carries about 4-6 bus loads of passengers to the city each morning.

      I suspect the timetable was developed in the days where most passengers caught the bus from Rosebank Road onwards — but the massive shift to trains in the MtAlbert-Kingsland corridor seems to have occurred so quickly the bus timetable has yet to keep up with this shift.

      Reducing the amount of buses on this route (to every 10mins during the peak and 15mins offpeak for now — increasing to every 15mins all day as more passengers shift to trains) would allow more buses to be operated on other routes which would benefit more from increased bus frequencies.

      • I agree James we do need to be smarter about how we use our bus resource. That’s a very interesting observation about New North Road – I have seen similar along there and Sandringham Road. However, be aware of whether the time you’re checking patronage is during university term time. Because New North Road buses run right through the university (whereas the train doesn’t serve the uni well) I’ve noticed that there’s a huge student use of those buses (which obviously largely disappears during non-term time).

  • When did the 40% tertiary discount come in? I think it could be responsible for the big March numbers in recent years.

  • The Trickster

    Also, re the decline of buses, when did the construction of the Central Connector start as that possibly would have had an impact on journey times resulting in mode shift.

  • axio

    I wonder if the upgraded CMJ had an impact. According to this was finished in 2006, although it is not clear when the key parts of it opened.

    • Matt L

      I was thinking the same thing, the CMJ improvements improved the motorways until induced demand kicked in a few years later and considering that most PT journeys are to the CBD then it might have made an impact. In reality it probably wasn’t one thing that caused it but a combination of things like better roads, cheaper petrol etc.

  • Sanctuary

    Mr. Arbury, I heard the mayor of Brisbane on the radio today saying eleven of their water ferry terminals had been washed away, at cost of three million each. This got me thinking about my favourite PT topic of water transport again. If I could be so demanding, could you see fit to a post one day on the pros and cons of water based PT? A glance at a map of Auckland makes a pretty good case that this type of PT could likely have great potential.

  • mark

    re the 2004-2005 drop – it may well be the “crash” in the foreign student sector. I can’t recall the exact period, but there was major drop due to exchange rate/security fears. Base students were arounf 20,000 – not sure of the drop – but was substantial for a couple of years.

    I remeber the story at the time was that the Dominion Rd bus route was hit with a big downturn – and it was put down to lack of forerign students.

    • The downturn in foreign students was certainly given at the time as the main reason for the 2005 patronage slump from memory. The industrial action and some big fare rises in 2004 contributed too.

  • Southernrata

    As a matter of interest, when did the Goldcard offpeak subsidy come in?

Leave a Reply