UN 196: FlexiRide - How productive is it really?

(scroll down if you just want to read about Tarneit) 

Why productivity matters

Economic efficiency and productivity matter. They affect whether we can have the things and services we want. Wealth accumulated from a period of growth does not imply wisdom to spend it beneficially, sustainably or fairly but at least gives you the choice. 

Also note the power of compounding, which Einstein described as the eighth wonder of the world. Small differences in growth rates between countries can compound to big differences over time. Consider this quote from US economist Tyler Cowen. 

Had America grown one percentage point less per year, between 1870 and 1990, the America of 1990 would be no richer than the Mexico of 1990. 

Australia, like many other countries, have had declining labour productivity growth in the last decade or so. Productivity growth doesn't automatically mean that the worker gets a fair share of it. But not having productivity growth makes it harder for economies to achieve a 'sweet spot' of low inflation, low unemployment and real wage rises. 

Australia went through a national conversation about productivity in the 1980s and 1990s. You had Paul Keating on the macroeconomic levers including dollar floating, wages policy, budget balancing and tariffs. Then bodies like the Productivity Commission lectured us about industry competitiveness and micro-economic reform. That is making various sectors of the economy more efficient with competition and choice (rather than planning and regulation) the preferred levers.

It wasn't all good; at their 1990s height the more extreme competition fiends had universal prescriptions that failed to account for differences between industries. Just as blunt as communists who saw state ownership and operation as a cure for everything. That's demonstrably untrue at least for public transport, with examples of both good and bad in both private and public operations readily available.  

High productivity for public transport is good..

Productivity is what you get out versus what you put in. A typical measure for public transport is passenger boardings per kilometres of service operated, though you could use other measures like passenger kilometres instead.

More people on public transport is generally a good thing, especially where service is good enough to be a popular alternative to car driving (and all the negative externalities that go with it). 

If you want public transport to meet more of peoples' needs for more trips then whatever you propose needs to be affordably feasible for the scale needed - not just for a few people or a small section of the city. As staff costs are such a high proportion of overall costs this means a high number of passengers carried per public transport employee. If you want the latter to be fairly paid then productivity is important.

.. but very high productivity might not be

If high productivity is good, shouldn't very high productivity be better? Right? Wrong!

Though in one sense a 'good problem', very high productivity is bad as it represents a network doing less than it should of the transport effort. This is because it signifies underinvestment where overcrowding on individual services and lower than possible network patronage both coexist. The latter also means low public transport mode share despite the crowded passenger experience. Add road congestion (as there are no good alternatives to driving) and you have a spiral where buses are unreliable (such as the notorious Route 525 in Donnybrook/Craigieburn and other routes in similarly underserved growth areas) and no transport mode in the area is good (a bit like Melbourne Airport but for different reasons). 

A solution is to invest in service hours and frequency even beyond the point where productivity peaks (but remains above average) as you are then maximising patronage. It's at this high patronage point, not the maximum productivity point, that the usage of and thus benefits of public transport are highest as explained here. Due to high fixed infrastructure and staffing costs this is even more so with rail; there you can simultaneously have crowded trains and a poor return on fixed infrastructure if frequency is kept low. The latter has been been especially so under the current state government that has fed infrastructure but starved service, with this declining per capita for metropolitan train and tram.  

There is a productivity 'sweet spot' in public transport. Too high is bad since it represents both overcrowding on individual services combined with lower than potential network usage. But too low is also bad since it limits community benefits from the network relative to what it pays. There is no denying that large parts of the transit network should be more productive than it is, with this being an essential planning and management focus.   

Institutional advocates for productivity in PT

Prescriptions for public transport reform from some 1990s era market economy evangelists seemed more ideological than pragmatic. And they didn't deliver all promised savings. For example the benefits and cost-savings of transport franchising were exaggerated with the first version of rail franchising needing a government bail-out. Though to be fair the advocates of state operation also over-state their case; there have been good and bad examples of both state and private operations in Australia. And sometimes it can be hard to fairly attribute cases of genuine productivity reform where this has been achieved within an organisation with little hoopla as it's just regarded as good day to day management. 

Failure is insufficient for an institutionally-supported intellectual or policy movement (like those who would have over-sold rail franchising in the '90s) to just die. Their competition-cheering ideological heirs remain in parts of government, large corporations, private consultancies, think tanks and bodies like Infrastructure Victoria.

They may still advocate prescriptions based on theory applied where it shouldn't. For example Infrastructure Victoria retains a mental model of public transport modes competing against one another. This doctrine was tried in the early 2000s but resulted in operators publishing fragmented information only showing (for example) maps of their half of the rail network. The failure of this was sufficiently apparent for the government to broker the Melbourne Passenger Growth Initiative which ultimately became Metlink with more integrated information published.

Of course this was before IV was established but this over-application of competition theory remains a doctrine in the blood of its staff as evidenced by their advocacy (sensibly ignored by the state government) for modally-based fares. Even though this is an example of competition undermining (rather than reinforcing) productivity because (i) one train driver can carry far more passengers at a faster speed than one bus driver and (ii) efficiency-enhancing bus reform is made politically harder if not impossible as some trips involving a change of modes become relatively dearer. 

So you can't rely on IV to be as consistently supportive of productivity in public transport as one might think, even though it has good ideas about bus reform and (likely) efficient infrastructure utilisation.

Similarly with DTP; the Bus Plan says the right things but implementation has been embarrassingly slow. DTP's internal processes have been a problem as has its difficulties in successfully advocating within government for it. Also, as mentioned later, parts of DTP have wasted time on forms of transport that are inherent low productivity, despite this being tested and known beforehand. Plus its leaders appear to lack performance incentives that might encourage productivity, such as a part of the Secretary's pay being dependent on mode share for public and active transport modes.     

Productivity successes

If you want reform and productivity on public transport it probably worth looking at successes both elsewhere and here. I'd nominate the mid-late 1990s and again about a decade later as being when the biggest recent patronage productivity gains in Melbourne public transport happened. I won't discuss V/Line much here except to say it's had strong recent growth, largely due to a tendency for the government to rely on it, rather than an extended electrified Metro network, for transport in the fast-growing west and north.  

The 1990s 'from a system to a service' reforms under minister Alan Brown cut staffing to save money. However reliability improved and at least some of the savings were returned to the system in the form of better service. Most notable of these were off-peak boosts (from 20 to 15 minutes) on the Frankston and Dandenong lines, the restoration of a full service on the Upfield line, and, biggest of all, the doubling of Sunday train and tram frequency between 10am and 7pm. Even though these were made during a time of relative parsimony in public finances, the relative size of these service upgrades has exceeded anything that has happened under the current government for metropolitan (but not regional) trains (and trams). This type of productivity gain was largely done by cutting costs, patronage rose but not very quickly. 

In contrast the 2000s increase was basically a result of surging patronage at a time where service growth was only a little better than flat for trains and trams. In a few years from about 2004 patronage was growing massively, aided by CBD development, high fuel prices and general population growth in Melbourne. Bus usage also grew greatly but in this case there was also a large service growth for a few years from the 2006 MOTC plan. Bus usage closely matched service increase. That's good by industry standards but it could have been higher if more bus network reform was done. 

Higher patronage is good for productivity because fixed costs are distributed over more passengers. But, as Melbourne found out, it can also lead to overcrowding and falling reliability unless other improvements are made. 

Desirable for rail include better maintenance, better scheduling, more off-peak frequency, higher capacity signalling and (on busy corridors) new tracks to reduce the knock-on effects of disruptions and increase capacity. Tram and bus would most benefit from on-road priority, network and timetable reform and larger vehicles (so one driver can carry more passengers). 

On the other side of the coin, if you want high productivity then you should avoid networks dominated by indirect bus routes with little unique coverage. Nuance here is important and productivity is best measured on a network rather than route basis. For example a two-tier network with four very direct, frequent and productive routes and two less direct, less productive routes may be more productive overall than a network with eight somewhat indirect and infrequent routes with only middling productivity. Such understanding can be lost if you try to split the network into profitable and non-profitable routes, such as was done in some countries where public transport was deregulated. 

Flexible route buses can be even less productive than even meandering fixed routes. How much less productive? Let's look at what's been stated to be the heaviest used flexible bus route in Melbourne. 

Tarneit FlexiRide case study

Starting in October 2022, the Tarneit FlexiRide commenced as techbro-inspired flexible transport mania was sweeping Department of Transport executives (remember the 'mobility as a service' hype?). DoT had developed FlexiRide as a new brand, taking over from Telebuses in some outer eastern suburbs. And it did seem to have a theoretical advantage in growth areas; a flexible route service with a small bus could be rolled out faster than a fixed route with stops, poles and timetables. 

FlexiRide is wonderful if you are the only passenger. Unfortunately it also maxes out if lots of people want to use it at once. Somewhere in between should be a 'sweet spot' where it can carry a reasonable number of people with reasonable reliability at an acceptable cost per passenger. Ideally this sweet spot should be over a wide range of usage numbers. But if it isn't then the scope for flexible bus routes to be affordably useful is narrow; perhaps only for less mobile people with special needs. 

Back in February 2023 the Wyndham Star Weekly carried an article on Tarneit FlexiRide, describing it as the most successful in the state. This had been achieved in its first 90 days of operation. 

But are FlexiRide's numbers actually that good and how do they compare with fixed routes in the area? 

The abovementioned article said that Tarneit FlexiRide was attracting 210 to 220 rides per day. The service runs about 15 hours per day (slightly more on weekdays, slightly less on weekends) so that's about 14 riders per hour. 

If that good or bad? We can't really tell unless we know how many buses are needed to run the FlexiRide service. 

Tarneit FlexiRide uses 18 seater Hino Poncho buses.  The fleet list for CDC's Wyndham depot show six buses with chassis dates in June or July 2022. Let us assume that one is a spare so five would be on the road, at least during peak times. One might also speculate that a smaller number than five is on the road outside peak times. If you were to spread that 14 riders per hour across (say) 4 buses then you'd get an average of around 4 boardings per bus per hour. Given that Tarneit FlexiRide is the most popular other FlexiRides are likely to be less productive. 

How does FlexiRide's performance compare with regular routes? 

My most recent data (obtained from DoT via request) is from Spring 2022. It shows that the quietest fixed route in Wyndham (153) had around 11 to 13 boardings per bus hour. Because Wyndham buses are such good patronage performers there's a lot of other routes getting 30 to 40 boardings per hour. Even on weekends.

In both cases FlexiRide is much less used. FlexiRide gets roughly one-third the passenger boardings of a quiet local fixed bus route. And compared to a busy route FlexiRide only attracts one-ninth the boardings. And bear in mind that both FlexiRide and regular bus routes require a driver, which is the most expensive part of running a bus service. 

OK, you might say that these are early figures and FlexiRide will grow with time. Maybe I should have given it a chance to grow rather than comparing it with established routes.

However in December 2022 I tried to plan peak trips using Tarneit FlexiRide. Even then it had maxed out, with the app suggesting a long walk to fixed route buses. 

This indicates that FlexiRide has already 'hit the wall' and prospects for further growth, at least in peak times are low. It appears to max out at a low level of productivity. That shows that FlexiRide is unsuitable in high travel demand/high growth areas like Tarneit and Melton.

The then DoT and the government were incorrect to go with FlexiRide (rather than originally proposed fixed routes) in 2022. However to their credit they've realised this and will be replacing Tarneit FlexiRide with fixed routes. Hopefully Melton, the second busiest FlexiRide, will soon follow.  

Conclusion

In public transport what doesn't scale doesn't count if you want to provide service to the most people for an affordable expenditure. That's especially so in growth areas where service provision often trails settlement.

If you want productivity then FlexiRide and like services are not your friends. But if your objectives are different, your target market is very small, you are not interested in connections to trains or other buses AND you are willing to tolerate high costs per passenger trip (possibly due to passengers having special needs) then flexible routes may be worth considering. But that's a different item to mass transit and is rarely a reliable and economical substitute. 
 

Click here for more Useful Network items