Jarrett Walker’s Human Transit blog has a post up about the study which compares Auckland to 13 other somewhat similar cities that I posted about yesterday. The post also usefully links to a full copy of the report. Jarrett helped put together the study, which compares many elements of Auckland’s public transport system to systems in Wellington, Brisbane, Perth, Adelaide, Sydney, Melbourne, Edmonton, Ottawa, Calgary, Vancouver, Honolulu, Portland and Seattle, coming to the general conclusion that on a lot of measures Auckland falls dead last.
The most common measure of “PT success”, and the measure which Auckland performs abysmally on, is that of per capita annual trips. That is, on average how many trips on buses, trains and ferries does each Aucklander take a year? On this measurement, Auckland has 44, with the next worst being Adelaide on 59, compared to the highest which is Ottawa on 168.
Jarrett’s post though, makes the good point that the usefulness of comparator city studies has its limited if the focus is solely on simple measurements like modeshare and PT trips per-capita:
Peer comparisons also carry the false assumption that everyone wants to be the same kind of city, and is therefore working to the same kind of goals. Low mode share for transit may mean your transit system is failing, but it may mean that it’s not trying for mode share, or at least that it has other objectives or constraints that prevent it from focusing on that goal. It may just mean that your city has different values. It may mean the city strikes a different balance between cycling, transit, and walking based on its own geography.
Helpfully this study goes much beyond the simple measurement, also looking at matters such as operating expenses, revenue per passenger, revenue per service kilometre, and service quality measurements. How Auckland scores on these measurements helps explain why we seem to do so poorly when it comes to per capita patronage. Service quantity and quality is low while fares are very high on a per kilometre basis.
While this information is all very interesting, I think that for comparator city studies like this to be truly useful, they need to be extended to look at two key additional matters:
- What benefits do cities with higher per capita patronage enjoy compared to cities like Auckland with very low per-capita patronage?
- What specific measures did successful cities take in the past that led to significant improvements (to patronage, cost-effectiveness, service quality etc.)?
I elaborate on the usefulness of the first point in my comment on Human Transit:
It would however be potentially even more useful for studies like this to go to the next level and start looking at the issue of “what does this actually mean for the city?” Per capita annual trips is a really good measure of a system’s quality, but what does that mean for the city?
For example, we know from this study that Ottawa has 168 trips per capita a year, compared to Auckland’s 44. What does this mean for Ottawa? Do residents spend far less of their money on transport, enabling a higher quality of living? Is much less of the city given over to providing road-space, allowing for greater development densities? Does the city have a more thriving and vibrant downtown – leading to better economic productivity? Does it have lower greenhouse gas emissions per capita? Has it been able to get away with spending far less on expensive road-based transport infrastructure over the past few decades? And so forth.
Politically it seems like the main justification for public transport investment is, at the moment, its ability to reduce peak hour traffic congestion. As I have calculated previously, it certainly does a pretty good job at this, but in order for a better understanding of the benefits generated by having 168 trips per capita per year, rather than 44, it would be great to dig a bit deeper into how cities like Ottawa really take advantage of the higher patronage statistics. Because, after all, high public transport use is the means to a better city, not the ends in itself.
Looking at the second issue above, the study doesn’t really look at this matter too much – although some factors of the Canadian cities (which perform incredibly well overall) are detailed, which probably provide us with a bit of an idea why they support so well:
Service quantity:
The quantity of service appears to be a significant factor in explaining why Canadian cities do better than US cities. Canadian cities have roughly 20-30% more service per capita, so this on its own should be likely to increase patronage, although perhaps not to a proportional extent.
Auckland, it should be noted, is in the same low range as the US cities in service quantity per capita and considerably below all the other Australian/NZ cities. This probably reflects decades in which public transport has been given a relatively low priority.
Wellington, by contrast, ranks near the top on service quantity, which probably reflects the city’s relatively long history as a dense and physically constrained centre with permanent public transport infrastructure (trolleybuses as well as rail). Wellington ranks generally among Canadian cities in the quantity of its service, though below the Canadian peers in patronage.
Urban form:
All four of the Canadian cities exhibit a high degree of centralisation of activity in the CBD and inner city. Calgary, Edmonton and Ottawa in particular all have strong CBDs and unusually high CBD densities given their locations. These high CBD densities have led to high parking prices by North American (though not Australasian) standards.
Vancouver’s CBD is famous for its exceptional concentration of high-rise residential accomodation at a range of price points. While the CBD does have a business district, as a whole the CBD has more residents than jobs, yielding a slight net out-commute. This unusual feature of Vancouver helps to provide for remarkably balanced loadings on major public transport corridors into and out of the CBD, yielding much better utilisation of service overall than would be expected in a single-centered city.
CBDs and other major centres rely on rapid transit, which means service running at high frequency and high speed, thus yielding high capacity. Rapid transit may be bus, rail, or ferry. What matters is not just the quantity of rapid transit, or the technology used to provide it, but how well it fits with the shape and demand-patterns of the city.
Beyond the CBD, the overall fit of development to public transport is important. Vancouver’s suburbs have built remarkable quantities of high-rise residential development immediately adjacent to rapid transit stations, generating permanent markets for these attractive services. In general, there is a high level of fit between the suburban development pattern and the rapid transit that serves it.
Finally, density overall tends to be slightly higher in Canadian cities than in comparable US ones, though density in the form that matters to public transport usage is impossible to summarise in a citywide statistic.
Major PT trip attractors:
Cities dominated by government employment usually achieve good public transport patronage. Governments are especially likely to encourage employees to use public transport, and tend to locate in centralised high density districts, usually parts of CBDs, where public transport is prominent. In Ottawa, the dominance of the Canadian government as an employer is obviously a factor in the strong performance, and this is also a factor in Wellington. Australian major cities are all state capitals, but most major US cities are not. Of the three US cities listed, only Honolulu is a state capital.
CBD-based universities also tend to be major generators of PT patronage. Cities with both universities and national/state governments located in the CBD, such as Melbourne and Ottawa, are likely to have an advantage on this score.
It would be particularly interesting to track this over time. To work out when patronage fell and rose in these cities and try to understand what was the cause for this change, particularly to work out what was successful in boosting PT usage.
Processing...
Also worth noting that measures such as PT trips per capita are greatly influenced by the choice of denominator, i.e. the unit of population. These units are most often based on jurisdictional boundaries, such as the Auckland Region, which may or may not be comparable between cities. There is no guarantee, for example, that the Canadian definition of Ottawa includes the same swathes of rural land that is included in our typical definition of Auckland
I think much more fine-grained analyses are needed to understand PT performance. This could be achieved by analysing the “density” of PT trips in smaller zones within cities. I.e. PT trips by census meshblock. You could then generate PT trips per capita rates for distance bands emanating from the city centre. Plotting these trips/capita rates against distance from city centre for the different cities shown above would be a useful extension to the analysis.
I suspect it would highlight that in central Auckland PT competes relatively well, but drops away rapidly as you get out into the suburbs.
The important thing is to compare like-with-like in terms of urban geographies:
* A city like Wellington, with a very high rate of bus use (>100 trips per person per year, according to Infratil) is not a fair comparison with the Auckland region as a whole; though it could be a fair comparison with the Auckland isthmus.
* Christchurch and the Auckland isthmus are also a useful comparison, as they have a similar population (360,000 v 450,000 for Auckland). As it happens, Christchurch and Canberra are very similar in terms of both bus use and population.
* The Auckland City region, Adelaide, Perth and Brisbane make for a useful comparison, because of their overall populations
* Sydney, Melbourne, Toronto and Montreal are another natural comparison as well.
Another comparison is Vancouver and Portland (mentioned in a discussion I was at yesterday). The two cities have similar population, similar geography, similar climate, and similar rates of car ownership – yet Vancouver’s rate of public transport use is double Portland’s, which is little better than Seattle’s, the city with which it frequently compares itself.
I suggested that this difference, in particular, had a lot to do with culture – not what makes people prefer one mode to another, but what makes people want to use public transport at all. Canadian cities have rates of bus use which are about two and a half times as strong as equivalent American cities – so it’s not the old LRT versus buses chestnut. Do other factors come to mind here?