IV's Bus Reform Community Research

 

Part of me thinks it's good that different parts of state government are talking about bus network reform and care enough to do research on it.

Another part of me thinks it's wasteful duplication, especially given that (a) both DTP and IV have been doing pretty similar activities (ie market research) on bus reform and (b) that it's all pretty academic given there's basically no state budget funding to do it.

As a rough simplification the Department of Transport and Planning (DTP) does stuff while Infrastructure Victoria (IV) is advisory. You might say that the government listens more to DTP than IV. However (unlike the big infrastructure construction authorities) DTP has a more subdued media and public image, with IV more active in media and public debate.

What recent work has DTP and IV done regarding bus reform? Here's a summary before I discuss what IV have most recently released. 

DTP

DoT's revived interest in bus network reform became public in June 2021 when the Bus Plan came out. Shortly after there were some small but worthwhile bus network reforms.

DoT then announced bus network reviews for a large slice of north and north-east Melbourne in September 2022. The review area goes from inner to middle to outer suburbs, with diverse demographics so could be considered reasonably representative of Melbourne. Over 3500 people took part in consultations in these areas (plus Mildura), though being opt-in and online the results won't be completely representative.

Questions included the trip purpose people used buses for, the times of day people travelled and the attributes people valued in a bus service (eg frequency, reliability and connectivity). There was also a lot of teasing out peoples' views on whether they would walk further to a bus that ran more frequently, or would accept changing between buses if it was possible to run routes more frequently. Questions were posed several ways. It seemed to me that they were skewed to win legitimacy for the department's preferred new approach to bus network planning and reform (ie having fewer but more direct and frequent routes).  

Read this for yourself in the consultation summaries published for each area here. 

IV

IV's 30 year strategy flagged bus network reform as being important. In December 2022, a little after the previous month's state election, IV released its own bus reform discussion paper which I reviewed here. The 'Get on Board' paper had a lot of good stuff analysing the current network and its issues. Many of those themes reflect what was in Victoria's Bus Plan that DoT released about 18 months previously.

Get on Board's two main weaknesses were (a) its misplaced faith in flexible route buses (that are neither productive nor efficient compared to fixed routes) and (b) their economists obsession with pricing over product, most corrosively in their backing of modal fares (that actually undermine efficient multimode networks where modes connect rather than duplicate). 

However the release of IV's paper led to media coverage and discussion on the role of buses. This was a good thing given that government support for upgraded and reform buses in the way that it counts (ie through its budget) is minimal, putting the service aspects of Bus Plan in a critical condition.  

Key DoT/DTP and IV developments regarding bus network reform are summarised below: 

IV's Bus Reform Community Research

This is the latest work to come out. Background on it here. Done by a commissioned market research company (Quantum Research), IV describe it as: one of the largest ever studies of current community perspectives on Melbourne’s bus network.

And it's true it does go into a lot of detail. Possibly too much. Because with this stuff you get to a point where once the basics are known anything further is gilding the lily and represents poor value for money. That's especially if (a) we already have bus reform examples to draw on including knowledge of what works and what doesn't, (b) it's research ground that others have covered before and (c) there's no significant funding to act on the research. 

Having said that we're always curious so it's still worth at least a flick through the report's 119 pages. It's dated March 2023, meaning it was finalised before the results came out from DoT/DTP bus review surveys. As you can see from the time-line above, the gestation period for the latter was unusually long (8 months). It is possible that there might have been some DTP/IV collaboration or at least cross-pollination. 

Let's look at the survey mechanics. The survey work ended in December 2022. It involved 4000 interviews. Pages 106 and 107 compare the representativeness of the survey sample with ABS population numbers. It was mostly representative except for a mismatch with regards to low income earners. For example ABS had 14% of households in 2021 with less than $26000 income versus the survey sample of 5% on less than $25000. Even noting the difference in incomes, this really under-represents lower income earners. This is significant as other parts of the survey indicate distinctive answers from this group with regards to bus usage. The unemployed were also under represented (2% vs 4%) but full-time workers were somewhat over-represented. Hopefully suitable weighting has been applied here. 

Pages 106 and 107 did not mention how ethnically/linguistically representative the survey sample was.  Whether one was from a culturally or linguistically diverse background had a significant bearing on some results, including propensity to ride buses frequently.

Things worth knowing

Frequent bus users are not that representative of the general Melbourne population. 16% of respondents used the bus at least weekly. The groups with the highest usage included those without car access (36%), 18-34s (25%) and those from a culturally or linguistically diverse background (24%). In other words roughly twice, 1.5 and 1.5 times likely respectively. Those who never used the bus accounted for 34%, with retirees and those in growth areas being about 10% above that. 

3 in 5 knew where their local bus stop went with only 1 in 5 knew where buses went. This gap could be a result of non-usage along with limited information and complex routes. Three quarters of people were within a 10 min walk of their nearest bus stop. 

Safety at night was a major theme, the latter especially amongst women. Those in growth areas and over 65 were also very concerned about this. This can have an effect on network design, especially the types of networks that force a lot of interchanging, such as the DTP survey led people to favouring. 

Many were willing to walk a bit further to a more frequent service. Bus users (unexpectedly) were happy to have more bus lanes for faster travel. There was more support for buses being removed from some streets if there was a flexible route service to replace it than if people were left with nothing. However flexible route buses are the sort of thing that sounds good but are both unreliable and have low productivity per passenger carried. It would be interesting if results would change if people were told these facts about them. 

Most respondents (70%) replied they couldn't survive without a car but a significant minority (25%) said they would like to get rid of their car if there was a viable alternative. So a car is thought as a necessary evil amongst some. But the view of buses isn't particularly high either, as later results reveal. Still it's possible for both perspectives to be prevalent and for buses to have a greater role than now. 

The choice model survey

Much of the survey consisted of people choosing between various options to make various trade-offs. For instance operating hours, frequencies, walking times to the stop and many others. Some options hardly exist now, eg 5 min frequencies. Others are unusual eg a given number of minutes early is equally weighted as that same number of minutes late when in real life earliness is far less accepted and tolerated. This asymmetry is why the industry has measures like 'no more than 5 min late or 60 seconds early'. 

For some reason the fare was always included in all options, which IV admits inflates its importance in the graph below. (click for a better view)

The steeper the line the more pronounced the influence. Below shows page 58 with the more influential factors according to the survey. Note the steepness of fares. Except potentially for short trips other passenger surveys I've seen tend to accord service-related factors like frequency and operating hours as being more important. Those not listed below (eg reliability and operating hours) were considered less important. And recall from before that the sample undercounts lower income people, so there might be representativeness issues. (click for a better view)

Also if you want to appeal to the 25% of people who may be open to not owning a car, then you do need wide bus operating hours. Even if such people don't use late night buses much, not owning a car is a strong factor that encourages peoples high bus usage at other times of the day. And the marginal cost of adding evening service is lower than (say) adding peak service. Therefore I would be wary about using this data as justification for cutting service levels at certain times unless there were some very strong 'greater good' reasons to do so. 

Something that came out strongly in the results is whether people would change to a connecting bus. This just had two 'required / not required' answers, with nearly everyone opting for the latter.

Yes we would all love a bus to our door. And if there wasn't then at least incredibly frequent services, eg every 5 minutes, that hardly exist on the network now.  Which by the way was the answer given to people in the survey, even if the frequency they were starting off with was a rather poor 30 or 60 minutes.

This hostility to changing buses is a different view to what some of the more radical bus reform advocates in DTP and academia would have preferred, and which came through in the former's survey. Answers might have been different if there a third option, eg a coordinated timed same-stop transfer for popular travel options, offered.

The responses here show the limitations of trying to design a bus network by opinion poll, that some trade-offs need to be made and you can't please everyone.  Transfer issues are only exacerbated if you include flexible route buses, due to their unreliability and ability to cover only a small area (unlike a long straight fixed route which at least can intersect with multiple routes, making many trips possible with just one change).  

In contrast to the negative reaction regarding transfers, if people were able to get the bus that went to their destination then they would be willing to walk further to catch it, especially if travel was overall quicker or there was higher frequency. The starting assumptions were 30 min frequencies on weekdays and 45 minutes on weekdays which isn't always representative of buses where survey respondents live. 

The survey includes a look at how different passenger groups responded to factors like service hours. As would be expected, inner city work commuters were least sensitive (possibly as many work in the CBD, often have trains and trams and mainly work M-F 9-5). Students in the western region were most responsive to increased weekend frequencies - possibly because of high needs, less rail infrastructure and relatively infrequent service existing. 

How much will you pay?

The IV survey was big in asking what people were willing to pay for (with the potential implication being that something is worthless if people don't tell a survey that they will pay for it). Survey respondents said they weren't very willing to pay for longer operating hours. They were somewhat more willing to pay for boosted frequencies on weekdays versus weekends, with their answers giving specific monetary values (ie a capacity to pay measure which would likely be wealth dependent).

Some of this might be influenced by existing services. Eg many outer areas have fairly low frequencies on both weekdays and weekends whereas parts of the inner west have good Saturday frequencies but poor Sunday frequencies. Other areas have fairly similar Saturday and Sunday frequencies but both are about 50% of that which runs on weekdays. 

I'm not sure about this approach. Having less money, the low income and unemployed respondents would be most unlikely to say they will pay much more for increased service on a per trip basis. 

What if you asked survey respondents whether they would ride the buses more if they ran longer hours or better frequencies? Here you might get more yesses. And they would likely contribute to more fare revenue to the system. So people would be paying more for PT per month because they are using and benefiting from it more, even if they answered no to a willingness to pay question. Plus there can be significant savings where improved public transport means that even a small number of taxi or uber trips can be avoided. 

Conclusion

IV has presented some substantial market research on bus reform.

Some of it rings a caution bell on matters like passengers willingness to transfer that might moderate the hubris of some. That's not necessarily a bad thing, with it important to heed lessons from Adelaide's failed bus reform attempt a few years ago. 

However it's also not a good idea to slavishly use its findings to avoid reform as you might not get a network much better than what we have now. For example you might have cheap modal fares, still mediocre service hours, slightly longer walks but still poor frequencies as there would still be lots of routes still needed to avoid transfers. And, due to that resistance to changing, along with cheaper bus-only fares, potentially also buses closely duplicating trains. In other words, especially if flexible route buses are part of the mix, a still inefficient and underperforming network. 

Sometimes you've got to ask people if that's what they really want, rather than relying too heavily on a survey, which doesn't necessarily explain all implications of choices made. 

Those seeking more information can watch a webinar on Tuesday 18 July. In the meantime please feel free to leave your thoughts in the comments below.