UN 192: What really happened to the Bus Reform Implementation Plan? (and more stories from PAEC)

In theory the Victorian Public Accounts and Estimates Committee (PAEC) is a way for the legislative branch of government (ie Parliament) to hold the executive branch (ie ministers and public service bosses) to account. A creature of the separation of powers doctrine generally common in parliamentary systems it does this by holding hearings, scrutinising budget papers, doing investigations and reviewing auditor general reports. 

In practice hearings may be dominated by (i) non-government members asking sometimes trivial 'gotcha' questions to make the government look bad and (ii) government members asking sympathetic 'Dorothy Dixers' to make the government look good.  Hearings are normally held around May (state budget time - budget estimates) and around November (for financial and performance outcomes). They are live streamed for those who want to watch. 

Despite sometimes inefficient use of committee time, PAEC hearings provide a means for ministers and top bureaucrats (eg DTP secretary Paul Younis) to answer questions on their responsibilities.  The committee comprises government and non-government MPs from both houses. 

Those appearing at PAEC hearings may get their staff to prepare some form of presentation or briefing with answers to likely questions. These can often be anticipated from media reports, known community concerns and previous questions asked in parliament. 

If those appearing don't know the answer to a question they can take it on notice, with a written response provided later. For example, here is DTP's written response to questions asked on the 20 November 2024 hearing. There's some interesting stuff there that I'll go through below. 

Bus Reform Implementation Plan

The Bus Reform Implementation Plan was meant to be the first action of Victoria's Bus Plan, released 1271 days ago today. Because the main bus plan was so lightweight (eg no specifics, no maps and no funding), the implementation plan would have provided critical substance.  

Aiv Puglielli MLC asked a pretty simple question on the status of the promised Bus Reform Implementation Plan. Other MPs (eg Trung Luu) previously asked similar but haven't had a good response. So it was a good question to bring up in PAEC. Here's the transcript (from p26 of DTP written response). 

 

You can see the responses from both Younis and Tieppo were unhelpful. They handball and delay on something they should have been able to answer on the spot. For this deliberately(?) obtuse dissembling duo (who rake in the big bucks courtesy of you and I the taxpayer) saying it's not done yet was apparently not an option. They didn't even say it was being worked on or that X, Y & Z had to happen first. 

The written response from DTP (received 28 November 2024) continues to dodge the question. It does not mention Bus Reform Implementation Plan (the document specifically asked about) once. Their preferred approach is to simply say what they've done (much of which is good, even if some numbers are optimistic and others poor measures, which I'll get to later). Read it below:  

The Department of Transport and Planning (DTP) is continuing to deliver bus network reform in line with Victoria’s Bus Plan.

Since the release of the Bus Plan, DTP has delivered new bus routes, simplified existing routes, and upgraded timetables across approximately 218 public transport and school transport services across Melbourne and regional Victoria, making the network simpler, faster and more reliable. We have introduced, modified and extended school bus services across the state, with a focus on growth areas. This has seen Victorians benefit from about 4,600 additional services running on the network each week.

The first Zero Emission Bus Franchises have been awarded, covering 131 public transport routes and more than 230 school bus services across Melbourne. Over time, these franchises will transition our bus fleet with more modern, zero emission buses, towards a cleaner, smarter fleet, and deliver better value for money from our bus network.

DTP is continuing to improve the passenger experience with trials of Rapid Running on Route 246, removing the need for a timetable and moving towards a headway based turn up and go service. All door boarding is reducing dwell times at stops to improve performance and ensure passengers get to their destination on time. Route network changes to remove duplication and simplify routes is making the bus network easier to understand, making the network more accessible to more people.

With over 4000 buses in the State’s bus fleet, progressive reform will take time, and ongoing reform will be required to respond to changes in technology, travel patterns and to integrate new communities, infrastructure and activity centres into the state’s bus network.

To summarise, mention 'Bus Reform Implementation Plan' and DTP leaders shift the discussion. 

Possibly partly because, GAIC growth area bus funding aside, DTP has difficulty convincing the government to fund service expansion (with the 2023 and 2024 budgets particularly disappointing). Secondly, with rare exceptions, there appears a reluctance to find savings to achieve a higher network aim, as required when doing bus reform in a budget constrained environment. Thirdly DTP lacks the growth mindset and marketing nous to sell the upgrades it does do. For example the wildly successful 7 day Route 800 bus was entirely marketed by operator and community efforts - not DTP's. Fourthly the department plays a weak defence game on revenue, with it even denying the extent of bus fare evasion. 

Without these four pretty basic competencies extending right up to the executive, deputy secretary and secretary levels of the Department you can't have a workable and funded Bus Reform Implementation Plan. It seems pretty clear that buses (and the missing Bus Reform Implementation Plan) aren't keeping the leadership group awake late at night, even though they should. 

Progress on bus priority

A bus carrying 40 or more people can easily be delayed by lines of cars carrying barely one each. Bus priority, including lanes and traffic signals can assist bus movement and improve reliability. Below, from Page 18 is an exchange between Aiv Puglielli MP, Paul Younis and Fiona Adamson (latter two from DTP):

Again a written response was furnished, as follows:  

 Priority bus lanes have been planned or implemented as part of a number of Big Build projects and major network change initiatives. These include Fitzsimons Lane, Mickleham Road, and Hoddle Street.

In the relevant period, the DTP Signal Network Optimisation Program, which reviews and optimises the metropolitan traffic signal network, has achieved the following related to buses:

• Bus improvements have been made to 731 traffic signals

• 148 traffic signal corridors (a route with multiple traffic signals) have been reviewed and 110 of those included bus improvements

Bus improvements typically consist of operational changes to the traffic signal system that reduce delays to buses.

It would have been good if the time savings were quantified. Eg did bus trips become faster, was variability in travel times reduced and how many bus timetables got rewritten with shorter run times (and ideally higher frequency) enabled by signal retimings?  

How much bus reform?

If forced to choose, I prefer lots of good stuff happening without much of a plan to the reverse of a great plan with little action. We know that the Bus Plan without the Implementation Plan is pretty skimpy but what about the evidence of action on bus improvement and reform?  

The answer you get depends on the measurement method used. I've discussed this several times. Here's a summary: (skip this bit if you just want more PAEC analysis) 

1. Counting annual service kilometres (and assessing trends, especially per-capita)

In March 2024 I discussed trends in public transport kilometres operated per year for each mode. I think this is a very good measure of service provision. Not accidentally it's also the one reported in budget papers so you can do long-term comparisons. It's a robust measure; if you do reforms like splitting routes the service kilometres does not change which is good. Dividing by the metropolitan or state population gets you per capita numbers. This gives insights like the per capita decline in metropolitan train and tram service and buses being about stagnant. Per-capita service trends can be compared between cities, as discussed here.  

Kilometres per year is what the performance indicator wonks call an 'output measure'. It doesn't measure network effectiveness, though you could bring in patronage numbers to get a productivity measure like passenger boardings per kilometre and check trends on that. A great deal of bus network reform, especially when resources are tight, is considering how best to deploy existing service kilometres most efficiently. Reforms may involve things like flattening peaks, straightening routes but then adding more trips, reducing route overlaps where usage is low, making spacing more even on overlapping routes you do wish to retain and more. 

2. Counting number of trips added

This method is fine if you are just comparing service on basically unchanged bus routes. For example you can say that Route 800 had X number of trips added in its recent weekend upgrade. It is also OK to express this as a percentage increase per day or week. Communicating number of trips added is easy and you get some big numbers if you annualise it.

As explained above, the minister has used this method, for instance "adding 20 000 new bus services in 9 years". Unfortunately without knowing the number of trips that already operate we don't know if that's big or small. 

Also, the method becomes misleading if you are reforming routes. You can even change trip counts even if timetables don't. For example if you split one long route into two shorter routes you have immediately doubled the number of trips operated without adding a single service. Conversely if you amalgamate routes you reduce the number of trips operated, but you aren't actually cutting service. Network reform can have similar effects. 

The wildly varying length of bus routes also causes problems, making comparisons meaningless. You can add very short shuttle routes with many trips. That disproportionately increases trip numbers even though it doesn't actually need many buses to run. Conversely adding long routes doesn't increase trip numbers very much, despite them being resource hungry to run. For these reasons I don't recommend this measure unless you are dealing with a simple upgrade to an unchanged route.  

3. Per route health check 

This is based on an itemised spreadsheet that classified each bus route according to whether it had major route issues (eg weak termini, indirectness, backtracking etc), major timetable issues (eg no Sunday service, frequencies worse than 'minimum standards', poor harmonisation with trains etc) or both. Checks could be done at various times to gauge progress. This was the first bus network health check, done when Victoria's Bus Plan turned 1000 days old in March 2024. It found that 33% of bus routes were healthy, without serious problems. 

I followed up with another health check in October 2024. This time 34% of bus routes were rated healthy. By this measure, which I think is pretty good, the current rate of bus network reform is slow. 

4. Bus service change events - activity comparison between cities 

Here I looked at a period of bus service change events for metropolitan bus routes in Melbourne and compared the numbers with Perth. The comparisons, by route, showed that Perth was doing about three times more bus service reform than Melbourne, or on a per capita basis, about 9 times more.

Changes were not weighted by their significance. That wasn't a problem in the Melbourne versus Perth comparison as I was satisfied that the mix wasn't wildly different, or if it was it was in Perth's favour. But otherwise not weighing is a limitation since different changes are differently important to the network. For example a minor rerouting or addition of a couple of trips on a quiet local route is less significant than a big frequency upgrade, a large extension or adding a new route, which typically requires new bus purchases unless there is substantial network reform. 

There is also no equality even if routes added are of a similar length. New school routes are very important as they save families time and may be the only form of transit in a growth area neighbourhood without its own high and sometimes even primary school.

But a new school route is typically one return trip per school day. Whereas even a basic public route may have 20 return trips per day and about 30 times the annual service kilometres per week counting weekends. The lesson is that if the government claims they've added, reviewed or reformed X new bus routes always ask how many are school routes as you may find it's a majority with much less effect on overall service kilometres than the impression conveyed. And you might uncover poor asset utilisation with many buses sitting idle most of the week when they could be carrying people on a public route.  

5. Chronology of achievements

I did this for Victoria's Bus Plan back in June. This gives a more qualitative impression of what's been achieved. Though unlike the annual service kilometres or health check methods you don't see its significance relative to the whole network. It is desirable to separate out major and minor initiatives. 

6. A network/service approach 

This broad method ignores individual routes and considers metrics like say (a) the proportion of the urbanised population within 400 metres of a service offering a minimum service standard 7 days/week or better and/or (b) the proportion of people within (say) 800 metres of frequent public transport.

DTP isn't very good at making what analysis it does here public. Plus the PPTN (which should guide any Bus Reform Implementation Plan) is in a poor state. But maps done by people like the Climate Council, SNAMUTS and Philip Mallis have filled this gap. Add the change in this map to the health check and per capita service kilometres and you start to get a good picture of the rate of network improvement and if it's keeping up with population growth.  

 
Getting back to PAEC, Page 22 has a request for the number of bus routes that have been simplified. The secretary said the number of routes was 'in the hundreds that we have viewed (sic)' and 'that there are still quite a few to do'. 

'In the hundreds' turned out to be 218 routes. The list is on pages 23 and 24 here. 

Not all those 218 are equal. Most are beyond scope if we're just interested in metropolitan public routes. Others, like the 27 changes to accommodate the rebuilt level-crossing free Croydon stations, represent activity but shouldn't count as real reforms. Unfortunately DTP's list counts them, giving the impression they are doing more bus network reform than they are. 

A slightly lower number ('nearly 200 routes either improved or added') was given by Mr Tieppo in response to a question from Michael Galea MP (see here for the transcript). Again it looks impressive but again counting number of bus routes isn't as good as totting up annual service kilometres added. And the bar for what counts as an 'improvement' can be quite low.  

A fairer indication of the extent of regular bus network reform in Melbourne requires a deeper dive. So I put DTP's data into a spreadsheet. Then I made pages that excluded school services, regional services and small changes like accommodating new station bus interchanges. The result of this is charted below: 

. 

DTP's numbers aren't 100% right as there are over and undercounts. Plus ambiguities like when you simplify 4 routes down to 2 remaining (as happened in Healesville) that gets counted as 4 routes reformed. Maybe that's OK. But I do know if you add 2 new routes and reform 2 others that would be counted as 4. The health check method, which looks at percentages, is possibly fairer here. 

I did note what could be omissions like the 279/293/907 upgrades & 603/604 downgrades of 2021, quantifying the Night Network reform of late 2021 and simplification splits like 380 and 834/835. Personally I'd have included them as they are some of the best (and certainly most cost-effective) bus network reforms DTP has done. However I've just gone with DTP numbers in the above pie chart, though the classifications between major and minor changes (which can be routes and/or timetables) are my own.  

Wearing our metropolitan route bus hat, this analysis cuts the number of reforms from DTP's claimed 218 to 57. Or 43 if we only count the more substantial timetable or route reforms. The more significant of this 43 include a large Craigieburn upgrade package (which saw most routes boosted to every 20 min weekday interpeak), some new routes including the 202 university shuttle and 475 and 501 in northern suburbs, 235, 237 and 605 Fishermans Bend enhancements, and more recently the 603, 604 and 605 reforms. Purists may say that basically just adding trips, like the 800 weekend upgrade, isn't real network reform but I've counted it since service uplifts like it (and even smaller ones) can still transform the network's usefulness.  

To test significance we need to know that Melbourne has about 350 regular bus routes. And that 43 that got done was over three and a bit years. So we're talking about reforms to about 14 public routes per year. When considered in the context of the abovementioned health check that still found issues with 66% or about 230 Melbourne bus routes, the pace of reform of the regular bus network is slow. Especially relative to the school bus network which has seen much more activity.  

This is a different impression to what DTP would want us (and PAEC) to believe. You could reasonably argue the toss regarding school and regional routes but the Secretary's counting of routes that are only changed to enter a new interchange as being 'upgraded and reviewed' is pure spin. Tieppo's 'nearly 200' is nearer the mark. However most work has been in the school route sector, with hardly a dent made into fixing the multi-decade reform backlog that continues to plague much of the regular metropolitan bus network or adding 7 day service kilometres in high-needs working-class areas like parts of Dandenong North, Thomastown and Campbellfield. 

Conclusion

Last month's PAEC hearing has educated us significantly on the changing status of Victoria's Bus Plan. 

The idea that it might have a written implementation plan attached appears all but dead, given repeated DTP failures to find funding, even if on the cheap through internal savings. But there's still been progress for buses, with some welcome though ad-hoc service upgrades and significant advocacy from the community. 

When these happen the government invariably describes them as being part of Victoria's Bus Plan, perhaps to save face. It is this action that is the true Bus Reform Implementation Plan; not a coherent document but whatever the government decides to fund for buses at the moment. And, to the extent that community bus advocacy campaigns are successful, it is these as much as DTP who could be said to have shared in its 'writing' as policy anarchy reigns.  


Other Building Melbourne's Useful Network items are here

[History] 15 years since the Metro Trains takeover

Today marks 15 years since metropolitan train operations changed from Connex to the Hong Kong-based MTR (trading in Melbourne as Metro Trains Melbourne or MTM).

My words and pictures from that first day in 2009 are here. The 2009 franchise, in particular, was controversial due to the rail network's deteriorating reliability since 2003. The widespread calls to 'sack Connex' were too loud for the government to ignore. Reliability eventually improved under MTM but not before the new more generous contract committed increased funding and the Brumby Labor government was defeated with rail services being a major issue in marginal seats.    

Keolis Downer, who operate Yarra Trams, will also be marking 15 years since they took over. However, unlike MTM their franchise will not be continuing. Instead, tomorrow they will be handing the keys over to Transdev John Holland.

This is the farewell video that Keolis Downer produced. 

The reason for the rail franchise also not finishing today was due to an extension that was given. This is to ensure continuity to bed down Metro Tunnel operations (with this starting next year). DTP will spend the next year or two working on the next franchise. 

A look inside Anzac Station

 Melbourne's Metro Tunnel is not yet open for service but I was fortunate enough to get on a tour of Anzac station on St Kilda Rd near the Shrine of Remembrance. This was offered to attendees of the Australasian Transport Research Forum conference as well as some others (like me). 

The station is still a construction site but is nearly finished. In this video I walk from the surface, down to the concourse, then down to the platforms, then back to the concourse then out again to the surface. This station will be a major interchange point with close integration between train, tram and a couple of bus routes. 

The Metro Tunnel is scheduled to open next year.

It will definitely greatly speed transport to the university and health precinct at Parkville as well as jobs and apartments around Domain. Other assured benefits include relief to crowded trams on the Swanston St corridor and increased potential to develop the Arden precinct due to its vastly improved accessibility. Improved reliability for northern group trains is likely as of the northern group only Craigieburn and Upfield lines remain in the City Loop, with busy Sunbury line trains now running via the Metro Tunnel.  

Potential wider Metro Tunnel project benefits include more services on lines like Craigieburn and Upfield (both of which have long suffered timetables that compare poorly with the all week frequent and marginal seated Frankston line) and substantial tram network reform. The latter being made possible by (a) the potential to move some routes from Swanston St to better served job-rich areas like the western part of the CBD (which is currently underserved) and (b) create short links to improve connectivity to emerging precincts just outside the Hoddle grid.

The realisation of such wider benefits (such as the business case relies on to make the project 'stack up') depends on the specifications of the timetable we'll get. These are not yet publicly known. Given the post-pandemic shift to working from home and the rise of weekend and evening CBD activity, off-peak service levels are as important as traditional weekday peak frequencies.

The Sydney Metro experience has shown this strongly with them now having a 5 min interpeak frequency. And Perth can claim a 7.5 min 7 day frequency at some of their inner stations.

Key issues for our Metro Tunnel and related timetables include:

(a) off-peak frequencies on the Metro Tunnel (a 10 min frequency is probably OK for stations like Watergardens and Dandenong but must be 5 min or better before considering significant tram reform; people won't go down to deep stations to face a 9 min wait for a one or two station trip);

(b) whether the abovementioned off-peak frequencies will apply at currently underserved times including evenings and Sunday mornings (where services have traditionally been inadequate to support major city activity and major events);

(c) the extent to which all day/all week frequency uplifts are confined to the Metro Tunnel lines or include others like Craigieburn, Sandringham and Upfield as assumed in the 2016 business case); and  

(d) the extent to which the timetable to be introduced for the Metro Tunnel sets off a process that revives the stalled 2012 Rail Network Development Plan to deliver all week 10 minute frequencies across the metropolitan rail network (starting with the very cheap to upgrade service to Ringwood and including other key lines like Mernda and Werribee). 

It is answers to questions like these that opposition MPs could have sought to get from DTP and government at the recent PAEC rather than playing 'gotchas' on train and network interoperability that was never a design intention.