How Brisbane’s car-free bridge changed the spread of the city

The main campus of the University of Queensland (UQ) in suburban St Lucia is enclosed on three sides by the Brisbane River. It provides an outstanding setting for arguably Australia’s prettiest university, but it significantly increases road travel times from the city centre and cuts the university off from Brisbane’s extensive southern suburbs.

Students and staff consequently mostly live north of the river, often at a considerable distance as the suburbs around the university are expensive. Until recently, those who lived to the south relied on small cross-river ferries or took a long detour by car or public transport to the nearest river crossings in the CBD (to the east) or Indooroopilly (to the west).

All that changed in 2006 with the opening of the Eleanor Schonell “green” bridge, which dramatically cut travel times from the south and to a lesser extent from the north. It’s known as the “green” bridge because cars can’t use it; it’s solely for “active” modes of transport i.e. buses, bicycles and pedestrians.

Fortunately, a number of academics from the School of Geography and Planning at UQ recognised construction of the bridge provided an opportunity to evaluate a ‘natural experiment’. They undertook a before-and-after study, looking at how the bridge affected student and staff decisions on travel and residential location.

Elin Charles-Edwards, Martin Bell and Jonathan Corcoran will publish their findings next month in the journal Urban Policy and Research (Greening the commute: Assessing the impact of the Eleanor Schonell “Green” bridge on travel to the University of Queensland, Australia). Gated, unfortunately.

They found the bridge had a dramatic impact on how members of the UQ community travelled to the St Lucia campus. In 2002, prior to construction, 39% of all students arrived on campus by car; however by 2011, after the bridge was built, the proportion plunged to 22%. For staff, the share of trips by car fell from 70% to 50% (see exhibit).

There was a corresponding big increase in the use of active modes over the period. The mode share of buses in particular jumped spectacularly; in the case of students from 27% to 53%. The increase for staff was even more remarkable; from 10% to 23%.

The key force driving these changes was improved access to the metropolitan bus network. The bridge lowered trip times by bus from the CBD and “opened up” new residential areas south of the river.

Students and staff took advantage of the improvement in accessibility. The proportion of all students living within a 10-minute bus journey of campus increased from 34% to 44% between 2003 and 2012; the proportion of staff increased from 21% to 37%.

The travel time to the university by bus from inner southern suburbs like Annerley and Wollongabba halved. The researchers say (1):

Prior to the construction of the bridge the campus was accessible from just seven Brisbane suburbs in under 10 minutes, only one of which (Chelmer) was south of the river. This increased to 17 suburbs south of the river (27 suburbs in total) following the opening of the bridge in 2006.

The number of students living in southern suburbs like Rocklea, Sunnybank and Cannon Hill increased markedly. However traditional student areas in the northern suburbs like Toowong and Auchenflower suffered losses; although their accessibility didn’t change because of the bridge, they now had to compete with “new” more affordable southern suburbs. (2)

The headline message from the Charles-Edwards et al study is that provision of good public transport infrastructure can lead to an extraordinarily large change in mode share. It’s also a striking example of how public transport infrastructure influences where people and activities choose to locate.

Further, US transport consultant, Jarrett Walker, says it shows that it’s not just rail that can shape the pattern “of real estate demand”; so can bus infrastructure (Brisbane: a city transformed by a bus link).

There are a couple of other important lessons from this project for public transport infrastructure planning.

One is that the success of public transport at UQ can’t be readily extrapolated to all other situations; this is to some degree a special case. UQ is a huge travel generator and it’s almost completely encircled by a wide river separating it from very desirable destinations a mere 200 metres away. This barrier wasn’t breached until 2006. As Jarrett Walker points out, “the issue here was classic chokepoint geography”. (3)

The other lesson is that the high mode share won by public transport at UQ is in large measure because cars got no benefit from the bridge. Had the new bridge also catered for them, it’s likely the change in mode share would’ve been very different. It reinforces the point that it’s rarely enough just to provide good infrastructure for active modes; a key requirement for significantly increasing their mode share is making cars less competitive.

The Eleanor Schonell “green” bridge shows that public transport can win spectacularly if it’s approached strategically. That’s a positive and valuable message; but whether it’s road, rail or bus, not all investments in transport infrastructure make good sense e.g. Melbourne’s proposed East West Link motorway and Doncaster rail projects. (4)

_________________________

1. I don’t think the travel times are meant to be taken literally; what matters is the relativities.

2. It’s not possible to say definitively that all the changes between 2002/03 and 2011/12 were due solely to the bridge. There were other factors in play; for example, buses might’ve become more attractive anyway due to progressive improvements in the bus system and cars might have been rendered less competitive because of increasing congestion.

3. There are a number of highly developed localities in Australia separated by major barriers, especially water, from most of their neighbours e.g. Balmain, New Farm.

4. The researchers say the mode share of both cycling and walking increased for staff over the period but fell a little for students (Table 3 in the paper). I don’t get that; unless staff out-number students, Table 2 in the paper seems to indicate the student mode share also increased.