Infrastructure Victoria paper champions bus reform

Infrastructure Victoria today released a discussion paper on bus network reform. Get on Board: Making the most of Melbourne's Buses says that with the right route design, cheaper fares and more frequent services, much more could be made of Melbourne’s extensive but underutilised bus network. Age report here.

IV's paper found that the existing bus network is unappealing, inefficient and uncompetitive with driving. This was due to indirect routes, poor operating hours and low frequencies. Reform to it is critical as buses are the nearest and often only public transport for 80% of Melburnians, especially in growth areas.

Inaction on buses will lead to road congestion, entrenched social inequality and higher transport emissions. This is a discussion paper. IV will publish a final report in late 2023 with the public being invited to have their say in the interim. They acknowledge the start that the government has made on this, with the release of Victoria's Bus Plan last year (evaluated by me here). 

None of the discussion paper's themes will be unfamiliar to Melbourne on Transit readers. Indeed the paper references my work in documenting previous efforts at bus reform such as Metplan (1988) and Meeting Our Transport Challenges (2006). But it's good that they are all in one place by an authoritative body.  

Existing bus network snapshot

The paper has much good information handy for those curious about issues facing Melbourne's bus network. 

For instance it explains how important bus access is to providing public transport for most Melburnians. This need will only increase as our suburbs grow. Despite the vastly greater reach of the bus network compared to trains and trams it gets the least use out of the three modes (20% of overall patronage, though likely more during COVID as bus usage held up better). 

Limitations with existing bus services with regards to operating hours, frequencies and reliability are also covered. There's some great graphs showing how much service levels fall off on weekends (see graph on p24). Maps show the main bus congestion hotspots. In short buses haven't kept up with modern living and travel patterns, a theme that I've emphasised on this blog. IV (correctly in my view) sees scope for unmet demand on weekends that could be realised if service levels were improved. 

Figure 21 (p40) has a map showing which routes meet MOTC service standards regarding hours and span. I think IV's judgement criteria is tougher than that of the planners who implemented service roll-outs. This is because many routes are not considered to qualify even though others would differ. Since the MOTC standards are a minimum safety net standard, I'd have liked IV to draw on other resources like SNAMUTS and my frequent network maps to highlight areas that have a 'good' (eg every 10 min) or at least 'satisfactory' (every 20 min) level of service based on a tighter frequency but looser walking distance criteria (eg within 800m). 

Page 33 has maps showing the extent to which buses and trains connect (on a Saturday). There's two main cases where connections are good. These are (a) where trains are frequent, eg every 10 min such as on the Frankston, Dandenong and Ringwood lines and (b) where trains may be less frequent but buses have been coordinated with them. The latter stand out in Brimbank and Cranbourne which both had relatively recent bus network reviews (2014 and 2016). 

Page 34 has a map showing low and high productivity bus routes with a 20 boardings per hour cut-off. Low productivity routes are most prevalent in parts of Melbourne's outer north-east, outer east and south. While 20 boardings per hour might be defined as low it's important to note that it is still vastly better than on-demand flexible routes whose network role is sometimes over-emphasised. 

Another caution I'd apply here is the need to think in terms of productive networks as opposed to just productive routes. If an area had (say) 3 popular direct bus routes attracting 30 boardings per hour but one less direct route attracting 15 boardings per hour as a 'mop up' coverage service then I wouldn't automatically be removing the quieter route without knowing more about its role. It may be that it's worth keeping as a reformed overall more productive network would have been politically impossible to implement without it. The network supporting role of these quieter routes should be acknowledged if there are popular direct routes nearby though this does not mean that they are not subject to efficiency improvements if these can be done without sacrificing significant coverage.    

SmartBus and the university shuttles are identified as higher performing routes. The former is an over-simplification. There are actually sections of SmartBus routes (especially some orbitals) that have their quite poor productivity concealed by only counting very long routes as a whole. Conversely there are non-SmartBus routes that have very high productivity, notably in high demand or higher needs outer suburbs like Tarneit, Craigieburn and around Dandenong. Some are the product of recent reform while others have been unloved for years. The discussion paper should really have highlighted these also as they too present opportunities for cost-effective upgrades.  

Mention is made that our bus network doesn't have much of a hierarchy in speed and service, though SmartBus and the university shuttles are elements of top tier routes. IV try to define a service hierarchy on page 17, based on rapid, connector and local groupings. It seeks to be consistent with DoT's classification for their bus plan which I critically reviewed here. I still think IV should have nominated specific operating hours and frequencies as part of their hierarchy groupings. My review of DoT's Bus Plan contains some suggestions for a simpler hierarchy that could apply across modes consistent with network planning work they did a decade earlier. 

Reform opportunities

Reform opportunities are divided into those that improve patronage, coverage, customer experience and financial competitiveness. 

Patronage improving reforms include network redesign featuring more direct routes with priority and improved speed along with longer operating hours and better frequencies. These are your basics without which you'll have a hard time getting any higher usage. 

The coverage list is very thin with just two main points. These include demand responsive buses, taxi subsidies and integrating land use and transport. Unfortunately the most important is unaccountably omitted - local bus network reform. This was a key theme of successful reviews in Wyndham and Brimbank. Because local fixed routes are often more productive than demand responsive options the result network is often more economical than if you had fixed routes providing patronage oriented service and all coverage service provided by demand responsive (like the University of Melbourne proposal for Melbourne's west). 

The customer experience measures are less important than the basics but are still nice to have.  Elements include quieter electric buses, all door boarding (good for speed too), prepaid fares (can be inconvenient too!), better live network information and improved bus stops, marketing and branding. I think there should have been better emphasis on multimodal network information, especially bus details at train stations. 

The fourth and final was financial competitiveness. Despite it being a major part of the recent state election campaign I think pricing is overestimated - driving is generally dearer than public transport already. It's far more important that a good service is provided. Unfortunately IV is still advocating modal fares despite issues I'll discuss later. 

Bus reform success and implementation

Where have we been successful? The 2014 Brimbank and 2015 Wyndham network reviews are rightly cited as examples. Both resulted in simpler networks, more direct routes and patronage gains. I do think further gains are possible with added service kilometres to boost the main routes to even higher frequencies over more of the week.  (Disclaimer: I was professionally involved in both). 

To its credit the government has announced bus reviews in Melbourne's north and north-east (along with Mildura). These are major bodies of work, each being far larger than Wyndham and Brimbank. 

My personal view is that bus reform should be implemented in a large number of manageable smaller changes involving (say) 3 to 6 nearby related routes each. I've generally made this my practice in the Building Melbourne's Useful Network series presented here. However the real detail we don't yet know about is the extent of resourcing for the extra 7 day service kilometres so important to delivering quality all-day service on the premium routes. That's important so we don't end up with a repeat of last time where many reviews were done but most recommendations were not implemented (which the IV paper mentions). 

The paper refers to bus reform examples outside Melbourne. I'd have liked greater mention of Perth, which I think has consistently shown the best bus reform culture in Australia over many years. Sydney's recent initiatives are also very commendable. Auckland however is mentioned. 

Bus network reform isn't all plain sailing. There are some precautionary stories where it's failed. I'm not sure that Canberra's experience was quite as bad as mentioned. But Adelaide's was. I wrote a detailed item on why that was so here. 

For a discussion you can watch IV's webinar, featuring various academics, practitioners and advocates here: https://www.infrastructurevictoria.com.au/2022/12/05/making-the-most-of-melbournes-buses/

IV and modal fares 

The main area where I think IV is wrong is with regards to its advocacy of modal fares. In their case advocating cheaper travel for buses than a similar distance trip if made by train. I believe this risks slowing the very service and network reform that IV claims to advocate. 

Also if it is politically impossible to raise train fares (note the 'race to the bottom' re fares in the recent 2022 state election campaign) the only response is to cut bus fares for an overall weaker revenue take. This imposes an opportunity cost when any available funding should have been used to make the network better rather than cheaper. 

Where does IV's modally fragmented thinking come from? Bodies like them (and IPART in NSW) appear to nurture a certain thread of economist thought inimical to integrated networks. They do not necessarily see the public transport network as a whole that needs to present a united front in order for it to be best equipped to offer a popular alternative to driving. 

Another bias of economists who dabble in public transport planning is a propensity to see too many issues as pricing rather than product (ie service) problems. Modes may be seen as stand-alone revenue and cost centres. There is sometimes a wish for them to be costed and priced differently because of variations in operating costs and a distrust of cross-subsidies between modes (since they make accounting harder, can conceal inefficiencies and could lead to suboptimal resource allocation). This view is different to that held by both passengers (who expect an integrated service) and best-practice transport authorities (who seek to provide it in all aspects including routes, timetables, information and fares). 

In Victoria this strand of thought was most influential in the 1990s when we broke up The Met and franchised our trains, trams and some buses. We even went a couple of steps towards disintegrated modal fares with Bayside train-only tickets and National Bus section fares presented as options to the multimode Metcard. 

We now know that franchising wasn't the money saver its proponents claimed (with it being almost exactly 20 years since the National Express Exit). And, until we came to our senses, we wasted untold millions on rebranding parts of the network every couple of years. Metlink was established to provide integrated  passenger information and single modal fares (as IV want now) were also  thankfully abolished. 

Modal fares actually undermine network reform and reduce public acceptance of it (which the discussion paper itself acknowledges is critical). An example will suffice. 

Suppose you had a bus network with routes that paralleled trains or trams. A network reform proposal might reduce network duplication and improve service frequencies with the trade-off being that you might have to change for some trips. Passengers would pay no more for this under our current fully integrated fare system. There may still be resistance but none will be based on fares. If resistance is purely based on the routes it may be possible to make a small change that makes a new network's implementation politically acceptable while still preserving nearly all of its intended benefits. 

However if fares for train and tram trips were higher than bus trips then passengers required to change would pay relatively more under the reformed network. Introducing a fare difference component would make getting public acceptance for network reform harder. 

Thus less reform would happen and we'd be stuck with more duplication, more complexity and less frequency across the network - all things that IV's discussion paper claims to be against. Other types of fare reform, such as reinstating fare zones, differential peak/off-peak pricing (also supported by IV) or adjusting concession entitlements, may have a stronger case if they harm neither network revenue nor modal integration.  

The role of flexible route buses

Demand-responsive/flexible route buses is an area where some in transport have been prone to accepting tech bro-inspired fads more than evidence. Page 35 of IV's paper has these service types occupying the least productive end of the spectrum. Even often criticised neighbourhood buses running infrequently on indirect routes typically attract more passenger boardings per hour (even if still less than IV's 20 per hour threshold). 

Low productivity basically means high costs per passenger. That's fine if you are willing to bear that cost as part of public transport's social role and your intended role is specialised (eg to help those with limited mobility access basic services). Unfortunately low productivity routes do not scale up if you want more than a handful of people using public transport (which you do). 

IV rightly mentions the poor record of many flexible route trials. To their credit they mention that any role is in low demand areas. However if we're talking about growth areas (where blocks are small, households are large and there is high work, education and other trip generation rates) then these are anything but low demand areas. The high patronage productivity performance of bus routes in growth areas like Tarneit and Craigieburn is evidence of this. You might have a flexible route in the early days of an estate's development but it wouldn't be long until it needs to be replaced by a fixed route. 

Page 66 suggests the following role for on-demand buses: 

On-demand buses offer a method to provide low demand areas a public transport option, instead of fixed route services. With this option, the planning of fixed route services could be more focused on patronage purposes, rather than on trying to imperfectly meet both patronage and coverage purposes.  

I regard this as a poor option. There is no reason why you cannot have a two tier network of fixed routes (such as in the 2014 and 2015 Brimbank and Wyndham networks) where more frequent and direct routes are focused on patronage with less direct neighbourhood routes fulfilling coverage aims. In this way you do not have to compromise the direct route network. As it happens even neighbourhood routes in these areas enjoy patronage productivities several times more than is possible with flexible routes. If coverage is needed add a fixed neighbourhood route. Only if for some reason this is impractical should one consider a flexible route. 

As economic wonks IV should be extremely interested in productivity as it provides a means to obtain maximum value for money from a bus network. Thus even by their own standards they should recommend flexible routes only as a last resort for niche markets.   

Conclusion

It's great that IV has done this paper. Bus reform is one of the top single things that could be done to greatly improve transport access for millions of Melburnians. If the paper is seen as a gentle needling of government (which should have done much more than it currently has) then that will be a good thing, especially given Melbourne's growth and bus reform's high cost-effectiveness.  

With many infrastructure projects now well underway and even drawing to completion in the next couple of years it's now time to think about how we can best maximise public benefit from new and existing assets including our train, tram and bus vehicles. Reimagining our buses (including a big boost in service hours) is one of the most potent (yet too often underappreciated) tools available. 

Comments are welcome and can be left below. You also have about 6 weeks to comment directly to IV on their paper. A webinar discussion is below.