Thursday, March 28, 2024

UN 171: New Route 475 - another weekday Useful Network addition

The latest 2022 state budget bus initiative to be funded started last Sunday, March 24. It is the new Route 475, a bus serving growth areas between Diggers Rest and Sunday. 

You might expect this to be the standard every 40 minute affair as we're so used to seeing with local buses. However you'd be wrong. And pleasantly surprised. Because not only will the 475 run every 20 minutes in the peaks but also weekday interpeak and early evenings too.

Operating hours meet and in some cases slightly exceed MOTC minimum standards for local routes. Weekend frequency is typically every 40 min during the day and 30 min at night, with both matching train frequency.  

Run time is approximately 23 minutes. On its own this isn't an efficient fit for a bus route that at most times runs either every 20 or 40 minutes. Fortunately the route has gone to the local Sunbury bus operator which runs other routes from the station. Hopefully it's been possible to find scheduling efficiencies through interlining with one or more of these routes. 

The 475's 20 minute interpeak frequency is actually better than the (Metro) train, which is currently every 40 min at both Diggers Rest and Sunbury. However I'm going to hazard a guess and say the latter is short-lived as Sunbury should get trains every 20 minutes or better when the Metro Tunnel  opens in 2025. 

Route number familiar? You're not mistaken. The previous 475 ran in the Keilor East area. It got replaced by the new route 469 and changes to the 476 in 2020. 

Passenger information and what this reveals

Firstly the good.

PTV were prompt to update its revised local area maps for Melton and Sunbury that feature the new route. PTV's website announcement presents a map of the route, such as one (should) see at bus stop timetables. Also good as it's not always done; PTV information work is sometimes done in silos with little use of material that may be produced elsewhere in the organisation despite potential usefulness for customers. Hence notices for bus changes at Yarra Valley and Donnybrook don't have maps on their main page, whereas those for Broadmeadows and Clarinda do.  

In 475's case those who write copy for PTV website didn't read (or lacked access to) the timetable. Hence when the 475 was advised it stated that it is 'every 20 minutes during peak hours on weekdays' when actual service is much better (extract below). 

People rely on DTP/PTV as a reliable source and widely repost its material. Thus if PTV gets it wrong misinformation can propagate.

Below is an excerpt from Sunbury MP Josh Bull's Facebook page about the 475 bus. Posted on March 26, the item perpetuates PTV's underselling with explicit mention of a 40 minute interpeak frequency.    

The above is no criticism of Mr Bull (who also happens to be Parliamentary Secretary for Transport). With information from a source that should be reliable he did his duty in informing constituents of an important new bus service. Unfortunately DTP too often lets MPs and other stakeholders down by providing incorrect or incomplete information.    

The above story is not a one-off omission to be casually shrugged off. It happens too many times for that. Rather it indicates an ingrained systemic tendency to not understand and then to undersell services on the part of PTV/DTP.

I attribute this to the following:   

a. Blindness to opportunity and the lack of a patronage growth mindset, despite usage being central to PT's community benefits, business cases for funding and even existence 
b. Retention of a commuter mentality that ignores the role of PT for the diverse trips that form the bulk of today's travel, despite the post-pandemic WFH revolution that has flattened travel patterns. 
c. A bias against buses. Because of its CBD location most DTP staff are inner suburbs dwellers who only use trains or trams.  
d. An unwillingness or inability to learn a product's benefits and features (eg flexible travel enabled by a frequent all-day timetable) when introducing it to the public 
e. Poor sales abilities and ability to inspire allies (including stakeholders like time-poor local MPs who are happy to sell initiatives that benefit their seat if presented the facts). In 475's case the MP is Josh Bull, the Parliamentary Secretary for Transport. 

All five are limiting beliefs that are restricting DPT's view of public transport, its potential for benefit and thus ability to drive growth. 

You can be sure that others in transport, whether it be Transurban, Melbourne Airport, the Level Crossing Removal Program or the Suburban Rail Loop Authority, do not box themselves in like DTP does. 

If DTP/PTV isn't getting into peoples' ears then others will be, with different agendas. It's perhaps no accident that if the government wants something big done in transport they will go to others (who think bigger and act faster) before the meek and constrained DTP. 

Wider implications for transport policy and priorities

What are some wider implications? Let's jump from one little bus route to the way tens of billions in transport funding is committed to in this state. Leadership might be a good place to start.   

The DPT Secretary's background is local government, not transport. This experience may be adequate to lead an administrative department but is perhaps less suited to be basically the network head. A role in which other cities might have the likes of an Andy Byford or even a Jeroen Weimar performing but which we effectively abolished when PTV was folded into DoT (later DPT). 

Why do I bring this up? Public transport exists in a competitive budgetary environment in which most proponents most of the time return home empty-handed. The government works within broad budgetary parameters such as income, expenditure and ability / willingness to borrow. That might have been forgotten in 2018 but is painfully obvious in 2024. 

Some expenditures are unavoidable and fixed while new initiatives are optional, more variable and often debated. There are many more requests for funding than will ever be granted. And some big commitments may put us on a path that may be favourable or unfavourable for other transport  initiatives. 

Aided by a supportive premier, transport infrastructure can claim to have done extraordinarily well in the last decade or so. 'Big Build' projects are typically overseen by separate bodies that are most certainly not apologetic about their existence or their projects' worth. Ken Mathers, for one, definitely wasn't. SRLA's Frankie Carroll or RPV's Nicole Stoddart regularly use LinkedIn to inspire staff and associates to convey a sense of joint mission. With a static page, DTP's Paul Younis does not use this channel despite its potential for good.   

Small infrastructure and service initiatives, despite high BCRs, have been much less funded. The much meeker DTP has carriage of these. As noted above it doesn't always sell the few service initiatives it does win funding for particularly well.

This occasionally frustrate ministers who've after fighting to get funding reasonably expect a publicity and ideally political dividend in return. The minister may put themselves out on a limb by publicly releasing plans (like for bus and tram), hoping the department's secretary would effectively advocate them. But if opportunities are not seized plans may fail to win funding and sink almost without trace (such as with the bus plan and possibly also the tram plan).

Whether you wish to sell something good you're doing, share a vision across your people and wider government or even get a minor bus timetable change done quicker than a level crossing removal, then DTP as it currently stands may not be the best equipped for the job. Other projects might cost more and have less merit but still win support with better organised backing.     

DTP institutional weakness risks creating a network comprising underutilised rail infrastructure and infrequent buses of limited usefulness due to the diverse travel patterns that modern cities demand. As we approach another time of budget stringency (partly due to high borrowings and subsequent interest payments for major infrastructure) a department that is almost apologetic for itself won't necessarily be the best equipped to win or even retain funding for the services a growing Melbourne needs. 

Update: Better late than never

The above regarding PTV underselling 475's off-peak frequency applied up to and after the service started running. It lingered long enough for it to be propagated by the local MP as recently as March 26. But, to its credit, PTV belatedly revised the wording on their website item (see below). 

This revised wording acknowledges the 20 minute all day frequency. And the reference to 'most of the daytime' is correct as there are some 23-24 minute gaps in the pm peak (defensible as pm peak train arrivals at Diggers are uneven).

However the reference to 40 minute evening frequencies still sells the service short as it is more like 30 minutes, especially on weekend evenings. That's an important distinction as many people would know that evening trains at Diggers operate every 30 minutes, making a timetable that remains coordinated until last bus a large user benefit.

While the new wording is much better than the old, there's still underselling. Hence the comments above about this being a feature of DTP's culture, at least with regards to buses, remain valid.

Given successive ministers' interest in promoting buses as a transport option, this is a restrictive mindset that they need to challenge the DTP Secretary to do better on.     


Route 475 is a welcome addition to the network in an area that hasn't seen changes to buses for years. 475's above average weekday frequency and operating hours should be attractive and ideally would set a precedent.

Earlier provision of buses in growth areas reduce cost of living pressures and contribute to long term patronage as fewer households will have needed to buy multiple cars even for basic trips.

The government has collected significant GAIC funds from developers of new estates. Hopefully some of these will find their way to funding the additional transport services so many of these areas need.

The weekday Melbourne Frequent Network Map has been updated to reflect this addition.

Other Useful Network items are here

Monday, March 25, 2024

Slow transport to fast cars: Learning from the Grand Prix weekend

In 1997 premier Jeff Kennett decided to privatise trains and trams after strikes disrupted public transport to that year's Formula 1 Grand Prix. Say what you like about him but that was how important he thought that working transport was for a government-defining event he snatched from Adelaide. 

In 2023 a state government-commissioned Ernst and Young report claimed that the Grand Prix delivered $268m in economic benefits, or a return of $2.66 per dollar spent, to the state (others disagree but we'll accept it for today's exercise). Gross attendance exceeded 444 000 people over the event's four days.  

Just last month Taylor Swift attracted 288 000 fans over three nights at Melbourne in her biggest ever shows. Public transport contributed hugely to the tour's success with thousands using the frequent trains added to the timetable. 

Then this last weekend was the 2024 Melbourne Formula 1 Grand Prix. They recommend use of public transport to get there, warning there was no public parking at the venue.  

Information focused on the 'last mile' transport to Albert Park from surrounding suburbs, with the pamphlet even titled Yarra Trams Travel Brochure (whose single mode title warns of what to come - but more on that later). This may be suitable for tourists staying in the CBD but not for those staying or living in most suburbs who might also have wished to attend.  

The Grand Prix is the sort of event (like football, cricket and concerts) that attracts occasional users to our public transport, including those who ordinarily drive for all their other trips. Melbourne generally has a good reputation for event transport that assists its standing as an event and sporting capital. Was this upheld during the weekend? Keep reading! 

Here's the PTV website item. It's dated 21 March so there might not have been very much notice given. It's also the first item that comes up when you do a Google search. 

When you start scrolling down you see this in relation to Thursday and Friday. There's a line about train services operating to normal daily timetables. Similar appears for metropolitan buses, regional (V/Line) trains and regional coaches. So there's a pattern - nothing more but nothing less either. Except for trams which do get a lot of changes. 

Say you're interested in the Saturday and Sunday. You keep scrolling down. This time there's no assurance regarding what's running. But the casual user might assume so given (i) what they read above for Thursday and Friday and (ii) it would be unthinkable not to run a full service on a major event weekend.  

By this time most will have (thought) they got all the information they need and won't scroll down. But if they did they'd see this unassuming little box near the bottom. 

Clicking on this gives a list of disruptions as long as your arm. However some would have missed seeing it despite its importance.  

Some disruptions are unavoidable features of long term projects. However the major Werribee and Sunbury line shutdowns (which made even Footscray dependent on buses) don't seem to be, with the works described as being for 'maintenance and renewal'. 

Shutting two lines reduces high capacity travel options and puts more pressure on rail substitute buses compared to if only one didn't run. The West Gate Tunnel road works wouldn't have helped either. 

And even if both lines were shut then there may have been a choice to bolster V/Line service (which normally has 40 and 60 minute weekend gaps from Wyndham Vale and Melton respectively) to supplement Werribee Line replacement buses. There may also have been scope to relax archaic V/Line rules about not allowing suburban passengers to board at stations like Sunshine and Footscray, at least for inbound trips. 

The potential for a frequent weekend V/Line service was demonstrated during last year's Avalon Air Show (48 000 attendees over three days) when a special timetable with intervals as close as 10 minutes was run. Such a frequent service from Wyndham Vale and Tarneit direct to Southern Cross would be superior to the Metro rail substitute offering that involved a train to Newport, slow bus (in traffic) to North Melbourne and an indirect (via City Loop) train to Southern Cross.

I received a report of such a trip (from Williams Landing to the Grand Prix) taking 133 minutes for a distance of little more than 20 km, indicating the substantial delays encountered. Even if people needed to take a bus to Tarneit time could have been saved, provided the service was frequent enough. 

The Grand Prix wasn't the only major event taking place over the weekend; there was also the Melbourne International Flower and Garden Show, making this last weekend a particularly poor choice to be doing works. 

Ten recommendations for major events transport

1. The minister seek a briefing from DTP, operators and event organisers on what worked well and what didn't for public transport in the last weekend. This briefing should include input from 'on the ground' staff and passengers.  

2. Have a state-significant events calendar on which preventable occupations (ie bus replacements) are not to be scheduled. A well-informed unit of DTP, working in conjunction with experienced transport operator personnel, Visit Victoria, the City of Melbourne and possibly the Premier's office to be the final arbiter.  

3. Simultaneous occupations on adjacent lines to be avoided unless absolutely necessary. Where occupations are necessary service levels on adjacent lines still running are to be increased, with passengers advised. Such adjacent lines could include nearby tram routes and V/Line if present.

4. Modified traffic control and dedicated lanes for replacement buses are a non-negotiable necessity for major occupations, especially for unpreventable ones on major event weekends. 

5. Pick up and set down restrictions on V/Line in the Metro area should be removed for the event duration and eventually permanently in directions and at times where there is no risk of longer distance passengers (who have fewer service choices) being unable to board.   

6. The planning of special event services should be done across all modes with visitors from a wide catchment in mind, and not just those staying in CBD hotels. The weekend's case where no new Metro or V/Line services were added is a poor example here. Ditto for passenger information material. 

7. The transport requirement of event organising staff (who need to be there earlier and finish later than attendees) needs to be a part of planning event transport services. Again it should be assumed they come from a wide area and use a variety of modes. 

8. 'Nasties' in timetables (eg 30 and 40 minute gaps in train and tram timetables and short operating hours for regular buses near the events) should be known so that extra trips can be added on major event weekends. While they may not be strictly necessary on capacity grounds they prevent bad experiences (eg lateness due to missing a Grand Prix Sunday morning train every 40 min or half hour waits at night) and have a relatively low marginal cost. 

9. Cutting maximum waits to make the regular network useful over more of the day and week. This would support major and minor events, regardless of whether services have been added or not. The ultimate aim could be Sydney-style 10-15 minute maximum waits, with the first steps being something like this: 
a. Cutting 40-70 min waits on the rail network to 30 minutes maximum. This is mostly Sunday mornings for Metro as well as weekends for Wyndham Vale and Melton V/Line. 

b. Ensuring 7 day service, longer hours and maximum 30 minute waits on key bus routes. 

c. Further cutting the 30 minute maximum rail network waits as achieved in (a) above to 20 minutes, starting with 7pm - 10pm evenings and 7am - 10am Sunday mornings. Followed by 10pm - midnight.  

d. Cutting 30 minute maximum waits on most tram routes to 20 minutes (mostly Sunday mornings and evenings) with similar for major bus routes (mostly evenings and weekends). 

The above would create a vastly more useful and connected network with maximum waits reduced from 40 to 30 and then 20 minutes across the week. This would apply even if rail does need to be replaced by buses as maximum waits would lower. 

10. A stronger operational voice within government that prioritises keeping services running (as opposed to replacing them with buses) with the spine to say no occasionally. It needs to be more widely understood that while often unavoidable, each occupation reduces passenger goodwill and, on major lines, undermines the case to use rail for freight by making it unreliable (with freight users complaining that they get even less consideration than passengers with regards to alternative arrangements).  

In our rush to build things, we also need to be mindful of the need to keep our city moving, including for public transport and major events. While there are new well-funded and effective bodies to push  forward with construction, the institutional voice for reliable operations, as one might have hoped to have been championed by DTP or a railways commissioner of old, has fallen silent or become ineffective. 

Disruptions are inevitable but were best possible endeavours taken to make transport to this year's Grand Prix as good as it can be? Possibly not. For example busy adjacent rail lines were shut down, no attempt was made to add V/Line service and on other lines extra trains were not scheduled, even where regular frequencies were as low as 40 minutes. A consistent failure to upgrade  cheap-to-boost regular frequencies, especially nights and Sunday mornings, entrenches long waits, increases crowding risks and reduces service legibility, especially for new and occasional users.  

We can and should do better if Melbourne is serious about maintaining its 'events capital' title.

If you travelled over the weekend feel free to share your experience below. 

Thursday, March 21, 2024

UN 170: PT service levels - Are we better served now?

Public transport infrastructure is worthless without adequate services to run on it. One should not get too far out of kilter with the other, otherwise either infrastructure utilisation or reliability will suffer. This should go without saying but hasn't in recent times.

The 'service first' remit of Melbourne on Transit requires me to be particularly sensitive to cases where (a) transit service lags infrastructure and (b) inadequate infrastructure seriously impedes frequent, reliable and connected services. This is a different emphasis to that of the current government whose 'infrastructure first' approach in Melbourne has consistently favoured infrastructure builds unbacked by service (as will later be demonstrated).   

Spiralling construction costs and budget-crippling interest payments could mean that the current government's borrow and build infrastructure agenda takes a breather. 2018 represented the high point when anything was possible, even trams to Rowville. The 2023 state budget started pruning and added little new. Expectations are even lower for this year's budget.

Changing post-pandemic travel patterns that flattened the peaks have made providing 'all day frequent service' a higher immediate network priority than peak capacity on most corridors. And community needs for such service are rising. For example largely migrant-driven population growth in unserved estates has rebounded, housing is under pressure and weekend Metro and V/Line patronage is booming. All this is putting pressure on various government services, not just transport. 

Campaigns, notably in Melbourne's north and west, are calling out inadequate train and bus services relative both to Sydney and more politically advantaged parts of Melbourne. So little has been achieved in the 1000 days since Victoria's Bus Plan came out that it can now fairly be called a flop after many meekly suspended judgment for so long. 

From the minister

Public and Active Transport Minister Gabrielle Williams is no doubt aware of this. On March 5 she sought to sell the government's record on public transport service. The key quote is below: 

This is not the first time the "20 000 new bus services" figure was used. On 24 October 2023, soon after gaining the portfolio, the minister told the Metropolitan Transport Forum that the government had added "20 000 new bus services in 9 years". 

This way of presenting achievements (by adding up numbers since you came to office) is exactly how you'd do it with capital projects like schools, hospitals and level crossing removals. Or job positions like teachers, police and nurses. 

Transport service additions is different. The number needs to be qualified on a per day or per week basis. Per week gives the bigger number so politicians tend to like it, as with this 2021 example for metropolitan trains. However it is also defensible because different days may get different numbers of trips added. 

Without qualification the claimed achievement becomes much less impressive. A single bus route with 8 trips per day is all that is needed to exceed the quoted '20 000 new bus services' (since 2014). That's a low bar and certainly incorrect. Yes, those unfamiliar with describing transport services can just as easily understate as overstate achievements. 

So how has public transport service really fared since 2014? To find out we need to delve deeper which is what I will do here. 

Counting trips - not always the right way

If you're talking about service upgrades on just one simple train, tram or bus line, it is reasonable to say number of trips. It's clear and people can see those extra entries in the timetable. 

Unfortunately, simply counting trips is a poor representation of service increase when applied across a network. This is because some routes are vastly longer than others. Adding a short university shuttle route (as the government has done several times) adds hundreds of trips per week. Whereas upgrading weekend frequency on (say) a long orbital SmartBus might require far fewer trips but cost much more. Similar applies when comparing metropolitan and regional train service additions. 

Service hours or kilometres operated per year are better measures of service resourcing that cancel out variations attributable to different route lengths. Budget papers use the latter so I will too. 

Just like one corrects for inflation when comparing peoples' wages over the years, one should account for population when comparing service trends. Especially in a rapidly growing city like Melbourne. With both annual public transport service kilometres and metropolitan population figures easy to get this gives a service per capita value that can reasonably be compared over time.

Previous service per capita measurements

Professor Graham Currie from Monash's Public Transport Research Group kept service per capita statistics for Melbourne up to 2019. Additional context appears in this 2014 engineers presentation

From about 2005 service grew across all modes in absolute terms. Metropolitan bus rose the fastest thanks to the 'Meeting our Transport Challenges' minimum service bus upgrades I recently discussed. Metropolitan rail was slow to grow but eventually took off on some lines, lagging the patronage boom by about three or four years (political context here). Tram usage also grew strongly but service hardly responded.  

There was a significant fall for bus around 2012. The 2014-15 Budget Paper 3, p247 attributes this to less dead running as part of the new metropolitan bus franchise. If so this should possibly not be counted as service. Thus I suggest caution in interpreting data around then including in the tables that follow. Especially as metropolitan bus greatly influences the total result. 

Absolute growth for all modes, especially train and bus, resumed the following year, albeit at a slower rate. But Melbourne's fast population growth exceeded service growth after 2012, with service per capita peaking in 2012 at just over 37 km per person per year across all modes. From then it steadily fell to about 33.4 km per person per year in 2017-18.

An updated version of the service per capita graph from Prof Currie had service per capita falling further to 32.7 km in 2018-19. This means that while it has added some public transport service, the Andrews government failed to arrest the per capita decline in its first term. In other words Melburnians had less service per capita in 2018 than they did in 2014. It is possible for the median Melburnian to have got more service but if more people were in growth areas with no or sparse service then mean service per capita could still decline. 

The cessation of this record series in 2019 has meant that we don't know service per capita during the pandemic and afterwards. There may even have been rises in the year or so population growth was lower. Election year 2018 was when its willingness to promise major infrastructure (including the Western Rail Plan, Rowville Tram and Suburban Rail Loop) peaked, while a few years previously, in 2015, was when state government interest in service reform started to cool.  

Since the minister was referring to their record since 2014 I want to go back further if possible. This is desirable because V/Line has emerged as a major rail carrier in suburban Melbourne and it would be worth seeing changes in its service levels. 

The budget papers

The state budget papers (No 3: Service Delivery) give aggregate service levels by public transport mode. This is in the form of million kilometres per financial year scheduled. Links to them here:  

2003-04 2004-05 2005-06 2006-07 2007-08 2008-09 2009-10 2010-11 2011-12 2012-13 2013-14 2014-15 2015-16 2016-17 2017-18 2018-19 2019-20 2020-21 2021-22 2022-23 2023-24

Budget Paper 3 information with regards to public transport service across the state (excluding school buses) is extracted below: 

All numbers are actual except for 2022-23 (expected) and 2023-24 (target). Metro total excludes Regional train/coach (i.e. V/Line) and Regional town bus. Notable data variations are said in the BP3 footnotes to be due to category changes for regional buses and coaches (2007 and 2008) and reduced dead running for Metropolitan Bus Franchise routes in 2012. The latter may mean that the indicated reduction in service may be less than appears. 

V/Line trains and coaches have had the biggest service growth. Annual kilometres operated more than doubled in 20 years, aided by infrastructure (RFR, RRL) that, unlike level crossing removals, had a large service component. Metropolitan bus, the main and often only public transport in growth area neighbourhoods, was most notable for its high service growth between 2005 and 2010.

Regional bus service kilometres is up by over 50% due to regional city network upgrades. In fourth spot is metropolitan trains. The main driver of service growth here has been electrification extensions in the north and frequency boosts in the east. Slowest growing are metropolitan trams. The stagnant mode in Melbourne's network, they have enjoyed few gains in network extent and frequency despite catchment densification and population growth.

The graph below summarises the above, showing service growth by mode relative to 2002. As for interpretation, if a line is at 1.5 this means that there's 1.5 times (ie 50% more) the service kilometres for that mode since 2002. 

Service per capita trends since 2002 

So far we've discussed absolute service levels. Even if annual service kilometres are maintained or even growing (which they have been) access to them may still be declining if the settled area is increasing (which it is). This is why service per capita is a good measure, especially for a city as fast growing as Melbourne is. 

The table below is the result of dividing annual service operated by June 30 population (ABS metropolitan or state figures depending on mode). For some reason the metropolitan totals are slightly higher than Prof Currie's but the trends are the same. 

Like Prof Currie I excluded V/Line from the metropolitan total, though it needs to be understood that in 2022-23 about 40% of its non-Southern Cross station boardings are now at metropolitan stations like Tarneit, Wyndham Vale, Melton and Deer Park. Not only that but the busier V/Line stations now get more boardings than most Metro stations who nevertheless enjoy between two and six times the daytime frequency, especially on weekends. 

In broad terms metropolitan service per capita rose and then fell while statewide service per capita largely held up, effectively preserving the 2000s gains. This is because (a) metropolitan Melbourne's population has grown faster than the state average and (b) the two regional service types grew faster than most metropolitan modes. Thus metropolitan service per capita is now very close to that for the state as a whole. Note that the 2021 per capita rise across Melbourne modes is a blip caused by the pandemic-related population loss (which was later reversed). 

The eagle-eyed will note that while metropolitan total per capita is just the sum of service km per capita on the three metropolitan modes, you can't simply add per capita service for all the modes for the Victorian total. This is because I'm using different populations based on who these modes predominantly serve. While not perfect, I've used metropolitan population for the three metropolitan modes, state population for V/Line and the population outside Melbourne (ie state - metropolitan) for Regional Bus. Adding per capita service of different modes also effectively gives different modes equal weighting when this is not so in relation to factors like catchment populations, network benefits, operating costs etc. Thus I'm reluctant to do it. 

Percentage change in service per capita

Specific service initiatives (or their lack) become more visible when you tabulate percentage change in service per capita by mode by year. Red is decline, orange is a small increase (up to 1%) while green is a larger increase.  

Note: Growth to 2023 not included as 2023-24 numbers are targets only. 2021 Melbourne increases due to the pandemic-related metropolitan population decline. 

It's easy to spot the times when governments were serious about service; just follow the green squares. In order of time these include regional city bus boosts (2005), regional train and coach (mostly Regional Fast Rail associated), metropolitan bus (MOTC program) and metropolitan train (driven by surging patronage 3-4 years prior). Priority then returned to regional train and coach due to the Regional Rail Link and Ballarat line upgrade projects a few years later. 2021's green squares are largely population related so should be ignored except for metropolitan bus which did see some real service growth. 

The graph below gives a wider long term view. V/Line is the big winner, with successive rounds of service upgrades that have enabled it to hold its position. Similar could be said, on a smaller scale, for regional buses. Also assisting the regional modes per capita performance is slower non-metropolitan population growth.  

Trams, despite densification in their catchments, have been neglected by successive governments, with 24% less service per capita than in 2002. Metro trains, the backbone of the network have remained stagnant per capita long term.  Metropolitan buses enjoyed gains in the first 10 years but have levelled out if not fallen per capita since. 

How has the current government performed on service?

With both absolute and per capita service information available, we can now discuss the government's record on service. As you can see below there are large differences between modes. 

The first point is there has been growth in the absolute level of service across all modes. When a minister speaks this is likely what they will point out. 

The second point is that the two heaviest used metropolitan modes (Metro train and metro tram) which account for maybe two-thirds of the state's transit usage, run fewer services per capita than they did in 2014.

Its reluctance to add service in a growing city means that the current state government is presided over a service level recession on the public transport modes Melburnians most often use. If one justification for this was equity, for instance by prioritising bus services in high needs areas, then this doesn't hold water given the failure of Victoria's 1000 day old Bus Plan.

Metropolitan buses may have just held their own on a service per capita basis, although even that is debatable giving lengthening multi-decade service backlogs in areas like Knox and the absence of any service growth in expanding areas like Pakenham, Officer, Mambourin and Mt Atkinson. A 
Metro Tunnel-associated timetable may boost service on some train lines but prospects for trams, in a city that professes to love them, remain distant.

The third point is that the lower patronage regional modes have done much better, with per capita service holding up. V/Line, for example, has had significant investment in service every few years since the early 2000s. Regional city buses have also had significant gains, though much of their higher per capita growth can be attributed to lower population growth outside Melbourne. Like with the growth area bus examples above, there is again a need for qualifications, eg the fast-growing Melton corridor remaining with an hourly weekend frequency for too long. 

Strong population growth, the low overall quantity of annual service kilometre increases and their deployment on modes with the fewest passengers may explain why governments and passengers may have differing views on public transport services and the extent of improvements made. 

The main caveat in all this is data quality, especially for the last few years. This is for both service (2021-22 being the last actual data) and population (due to the decline and then the rapid Melbourne rebound). The release of another year's of service data in the May budget and population from ABS later this month will help clean much of this up. 

How far is infrastructure ahead of service?

There is no doubt that Melbourne's trains and buses are what a private business would call 'lazy assets'. That is they sit idle for many hours of the week when they could be in revenue service (and attract reasonable patronage). 

The graph below compared trains per hour frequency of three major Melbourne lines of roughly comparable length across most times of days. All lines operate a frequent peak service with at least 8 trains per hour average arriving at Flinders St between 7:00 and 8:59am. Not all trains serve all stations but gaps are not serious. In contrast Craigieburn and Mernda lines operate at only half Frankston's frequencies at most off-peak times. This is despite potential to be much more frequent as demonstrated by the intensive peak service that operates. 

In more passenger friendly terms 6 trains per hour is service every 10 minutes, 3 trains per hour every 20 minutes, 2 trains per hour every 30 minutes and 1.5 trains per hour an embarrassing every 40 minutes.

Below I take the am peak trains per hour and assign them as 100%. Then I compare off-peak trains per hour for each line and compare them with peak frequency. The Frankston line, with its 5 minute peak/10 minute off-peak service, holds up better in all time slots. Meanwhile service collapses on the busy Craigieburn and Mernda lines to between 18 and 38% of peak service. This is despite catchment demographics that lend themselves to high all-day usage and higher observed loadings on the off-peak services that do run relative to the Frankston line.     

This exercise shows that public transport service badly lags available infrastructure, with the difference often greatest on lines with high social needs that have reliably voted Labor. While I covered trains the same applies to buses, especially on weekends, as I discussed here. The numbers demonstrate that the current government has let service stagnate with the result that it now has a lot of underutilised assets that the community is not getting full value from. 

What if service per capita had held up?

What if the Andrews-Allan government had built a little less and used the money saved to retain service per capita at the level it inherited? What sort of gains would have been possible? 

For Metro Trains that would have meant we'd have retained 5.0 km of service per capita rather than the current 4.8km. That's 4% more. Add that to the current 24.9 for about 1 million more service km per year or 20 000 per week.

If applied to ten 40km long lines that's 50 extra trips per week or 7 trips per day. That should be enough to cut maximum waits from 30-40 to 20 minutes across almost the whole network between 6am and almost 10pm all week. This would be a major advance for the small increment in service needed.

And it would set the scene for a staged roll-out of the minister's much mentioned but little progressed turn-up-and-go 10 minute frequencies, with a Ringwood upgrade being especially cheap (with marginal seat benefits too). Right now all eyes are on seeing whether the proposed timetables associated with the Metro Tunnel reverses the per capita suburban rail service drop, and whether all lines stand to gain or just a few.  

Retaining 2014's per capita service would have benefited trams even more. And they deserve it due to their catchments densifying. By now we'd likely have eliminated 30 minute Sunday waits and got evenings from every 20-30 to every 15 minutes all week on a lot of routes. Which is progress towards a 10 minute service (especially since so many are almost there at every 12 minutes). 

Buses sit at about their 2014 per capita service. But if you consider that buses were underserved in 2014 and populations in outer areas have grown faster than the metropolitan average one should (along with suburban V/Line) use more nuanced population samples to reflect local growth and needs. If this was done Pakenham and Melton would stand out as two areas with large per capita drops in service.  


More trips run on our public transport than they did in 2014. However service levels on the busiest metropolitan modes have lagged population growth. The result is a per capita service recession on the busiest metropolitan modes that leaves thousands more people underserved each month.

Millions more would have had access to better public transport services had the government just retained service per capita at 2014 levels. But they didn't for metropolitan train and tram while being sluggish on buses. 

Will the ministers and government now act, even if they might need to slow (now arguably less urgent) infrastructure builds to fund it? 

Saturday, March 09, 2024

How healthy are our bus routes? (1000 day Bus Plan special)

Today marks 1000 days since Victoria's Bus Plan was launched. It ably diagnosed what was wrong with buses and outlined what improved services might look like. But it lacked specifics on what was to be done. That was meant to be covered in a later Bus Reform Improvement Plan that we all politely waited for. 

After an initially promising start in 2021 and further wins thanks to the 2022 state budget a large scale bus network review was announced for Melbourne's north and north-east two months before the state election.

However momentum had waned by early 2023 with warnings and then the reality of a tough 2023 state budget with very little new for buses. This caused me to query the health of the Bus Plan in June 2023 with a grim prognosis, even as others like Infrastructure Victoria, the Committee for Melbourne and Friends of the Earth were stepping up their bus research and advocacy.  

2023's end saw no apparent revival. It was then 931 days since the release of the Bus Plan. Not even the Bus Reform Implementation Plan had come out. So I devised an online clock to see if we'd see movement within 1000 days. That raised significant interest, with '1000 days since the bus plan' quoted by campaigns and in parliament. Which is actually today. 

To mark the occasion I did a desktop health check of all 349 regular (ie non Night Network) bus routes in Melbourne. Such a check, accompanied by punctuality data, patronage data and much more, would be a necessary prelude to any implementation plan. So what you read here is just a start. 


The check is crude. It's based on existing routes, not peoples needs. There are just two measures: timetable and route. Both are judged on a yes/no basis with the result added. Thus the only possible scores for a route are 0, 50 or 100% even though it's more nuanced for each route. Still, adding the results of 349 quick tests is enough to convey a fair picture of the bus network's health which is what we want today.  

Timetables are mainly assessed by whether they meet 2006's minimum service standards. That is 7 day service at least hourly until 9pm, with 6, 8 and 9am starts for weekdays, Saturdays and Sundays respectively. Not a high bar but it is a service standard on which significant progress was made (though not recently).

I also compared timetables across days of the week, especially for busier or main road routes. If a major shopping centre route ran every 15-20 minutes Monday to Saturday but dropped to hourly on Sundays then I would mark it down as a major issue. As I would for a route whose frequency was irregular, was unharmonised with trains or over-serviced for its catchment. All of these would count as a major timetable issue.

On the other hand a peak, university, industrial or limited shopper route that lacked weekend service would not be marked down given their peripheral network role. Neither did I much consider overcrowding, appropriateness of run times or punctuality (though you can check the latter here). So don't be too surprised if I didn't pick up a timetable issue on routes you know there's problems with.   

Routes were rated on many factors including excessive indirectness (some being necessary for coverage), complex loops, deviations, weak termini and duplication with other routes. Minor problems were overlooked but your judgement may differ from mine in a particular instance. Overall I've erred on the 'soft' side; like with timetables there will be routes with issues that my desktop check still gave a 100% rating to (when 70% might be fairer). 

Still, this exercise should give an idea of whether bus reform is justified and an idea of its potential gains.  

Data summary

One third (116) of Melbourne's 349 bus routes had no major timetable or route alignment issues identified. The remaining two-thirds did.

Of those about 40% (ie 94 or 27% of the total) had both timetable and route issues. Especially where routes can be made more direct or duplication lessened this presents an opportunity for cost-effective  timetable upgrades with freed-up service kilometres.   

67 routes (or just under 1/5) had timetable issues only. Fixes could be anything from extending hours to minimum standards, adding weekend service or harmonising headways with trains for improved connectivity. While some funding is likely needed for the extra drivers and route kilometres timetable only upgrades are relatively easy with no public consultation or even extra buses needed (if done at off-peak times). Occasionally the problem is overservicing with a potential to transfer service kilometres to routes or time periods that need them more. 

72, the remaining fifth, largely need route reforms. Examples include removing a deviation, making service more direct, removing duplication or extending to a logical terminus like a nearby station or shopping centre. Cost-effective opportunities for improved service may be possible where multiple routes inefficiently overlap. 

Service upgrades over time

In November 2008 I checked the progress of the Meeting Our Transport Challenges program of minimum service standards for buses released in mid-2006. There had been some Sunday service additions in 2002 but evening service to 9pm remained rare in 2006, with only 13% of routes having it. MOTC upgrades had more than tripled this to 44% of routes in less than 3 years. This growth from 40 to 137 routes means an average of 50 routes per year gained minimum service standards during this period. For context Melbourne's metropolitan population was just under 4 million in late 2008. 

Where are we over 15 years later? We now have about 1.3 million more people and 40 more bus routes (rising from 309 to 349). The proportion of that 349 meeting minimum standards (including 7 day service to 9pm) stands at 61% (ie 213 routes). Or 63% if we are generous by discounting weekday only peak and university routes that you'd never run weekends. 

The gain from 137 to 213 (ie 76 routes) represents just 5 routes per year gaining minimum service standards between 2008 and 2024. In other words the rate that Melbourne upgraded bus service slowed by about 90% compared to 15 years ago.

Note: Contains only one intermediate data point. Thus the 2008 - 2024 trend will vary in rate. 
For example there were significant improvements in 2009-2010. And others in 2013-2016. 

Annual scheduled service kilometres is even better in that it counts all bus service initiatives. The 2015-16 budget papers reported that metropolitan buses had 110.8 million km per annum projected to be scheduled at the end of the 2014-2015 financial year (ie before the current government was elected). The 2023-24 budget papers had 129.3 million km per annum as a target for that financial year. This  does represent service growth in absolute terms.

However Melbourne's population grew even faster, by nearly a million, over that time (from 4.3 to 5.2 million). That means there is less bus service per capita now than in 2014. That would put Melbourne in a per capita bus service recession. Similar comments likely apply for metropolitan train or tram but not V/Line whose services have grown the fastest of all modes.      

It is in this less than buoyant context that recent statements from the Minister for Public and Active Transport with regards to her government's record on public transport service should perhaps be viewed. 

Some more numbers

Below is a table I've made of the raw spreadsheet data (which you can download here). I've already discussed the first data column re MOTC minimum standards compliance. N/A means routes like university shuttles that I excluded from the survey as they are not expected to run 7 days. 

There's a stack of routes that run 7 days but don't meet MOTC standards. This is mainly because they start too late or finish too early to qualify, especially on weekends. These would be very cheap and beneficial upgrades, especially where they include popular but underserved routes like the 630 on North Rd. 

While one could argue that not all quieter routes should run 7 days or MOTC hours, there's enough 'have not' routes (124) to demonstrate that a large number (at least half) should get upgrades on pretty solid patronage or social need grounds. Examples include key routes like 237, 281, 284, 404, 414, 468, 503, 506, 536, 546, 548, 549, 612, 800, 802, 804, 814, 844, 885 etc. 

The third column deals with timetable issues. Nearly half the timetables were identified as such. As noted before these might include not only non-adherence to minimum service standards but also other factors like headways unharmonised with trains or a big drop-off in weekend (especially Sunday) frequency versus other days. This (and the next) column is less objective than the first two so your judgement on this will vary from mine. This column forms half the final score. 

The last column on the routing forms the other half of the score. Again I saw serious issues in the alignment of nearly half Melbourne's bus routes. This includes cases where the route is sound but there's enough duplication with others to query the network's efficiency.   


The above charts and tables are based on spreadsheet data that you can download below. 

This includes comments for most routes to justify their scoring. 


This desktop review of all Melbourne's 349 regular bus routes shows a strong need for bus network reform. Most routes had issues with their timetables, alignment or both.

In addition the data presented indicates the extent to which 7 day bus upgrades have almost stalled since the MOTC program despite our city adding 1.3 million people in the interim. We've added service but  it's not been enough to keep up with population.

A failure to add sufficient service kilometres also harms the prospects for the sort of network reform envisaged in Victoria's Bus Plan. A major lesson from Auckland's success is that bus network reform is harder without extra service kilometres because you can't limit political risk by, for example, retaining some less direct / less frequent coverage style routes in high needs areas.

That doesn't mean you should give up on bus service reform. To the contrary. The more you pay attention to duplicative or overserviced routes the more 'greater good' improvements you'll be able to do. I described this more in Bus upgrades for a broke government. And if you've got processes to make several timetable adjustments in a year (rather than it taking it several years for one) you can follow the steady Perth approach of eventually getting what you want (ie simpler frequent routes) with a minimum of political backlash. 

Opportunities are especially high due to Melbourne's bus reform backlog. That can be demonstrated by the continued existence of embarrassments like routes 558, 566, 624 and 800 with severe alignment and/or timetable problems persisting for decades. Yet the record points to a reform stalemate, with the rate of even simple timetable optimisation exercises lagging cities like Perth.

It can't all be down to money; organisation and efficiency play a part too. Not only does Perth have more effective institutional frameworks and political support for improved buses but they also tolerate operational inefficiency and fare revenue loss less. Despite DTP having had a dedicated bus reform team there is as yet no sign of the promised Bus Reform Implementation Plan (that on March 6 was asked about in parliament). And, despite warnings, DTP has pursued failures like FlexiRide that have increased rather than decreased the cost per passenger carried, especially in high patronage areas like Tarneit where services commonly max out

Having now entered the Victorian Bus Plan's second thousand days, we hope that achievements in this period will greatly reverse the stagnation of its first three years and the per-capita service fall going back even further.

With a tough state budget mooted, network and service reform is perhaps the last and so far largely untapped hope for cost-effective improvements in public transport. However for this to happen political will and delivery capacity both need to be vastly better than now.

The ball is now in your court ministers Pearson and Williams.