Saturday, December 15, 2007

Route 465: Melbourne's most successful rail feeder bus?

While it lacks the profile of some longer routes or the gizmos of a SmartBus, Route 465, operating between Keilor Park and Essendon could well be our most successful rail feeder route.

How can this apparently ordinary middle-suburban route do so well?

I can think of four reasons.

Firstly it is direct. 465 runs straight along Buckley Street almost to the end. Then it turns north along Milleara Drive before ending at Keilor Park.

Then there's the trip generators. Unlike any SmartBus route, the 465 serves no regional shopping centres or university campus. However Essendon, like Hawthorn and Mentone, is known for its range of schools. While they are not all on the route, most are less than 10 minutes walk away. There are other routes in the area but 465 is the most frequent and direct, so it would take a good share of this traffic.

The timetable is also very significant. Starting before 6am, peak service levels are the best of any bus route in Melbourne. 7 or 8 buses per hour run in the morning peak which is about the same as trains from Essendon Station. In only a few cases do intervals exceed 10 minutes, with some as short as 5 minutes. Both peak train and bus headways are irregular but service frequency on both makes exact harmonisation unnecessary and there is a welcome absence of 'exact time' arrivals.

Off-peak times see a 20 minute service on weekdays and Saturdays. Sundays receive a 40 minute service, matching every second train. While one might quibble at the Sunday service offered, it is at least headway harmonised, and there seems to be less Sunday trading in the northern and western suburbs than in the inner, bayside and eastern suburbs. Services finish early at night but train/bus connections for those that run are generally excellent.

As well as the generally train-matching service level, the fourth factor is information and connectivity. Train times are usefully listed in the timetable so passengers can gauge transfer times. Typically this is between 3 and 9 minutes and somewhat longer on Sundays. Given the 20 minute train headway, and a 10 minute average wait if no effort was made at connections, the 465 rates fairly well for connectivity in the timetable. Adding to this are the times when drivers wait (within reason) for late-running trains.

Leaving aside its finishing times, little old 465 is a genuinely smart bus in many ways. It could well be an example for other routes, especially where stations are experiencing parking pressure (peak periods) and connectivity is poor (off-peaks).

Monday, December 03, 2007

Improving access to suburban stations

With our current suburban rail network (and any conceivable extensions to it) the majority of our city's population will always be beyond reasonable walking distance from a station.

The question then is how do we bring fast public transport within reach of more people.

Possible solutions to consider include:

(i) More walkable streets for faster, more direct and safer pedestrian access
(ii) More cycleways and bicycle shelters at stations
(iii) Better feeder bus services with improved connections
(iv) Park and Ride

The first two will appeal to some passengers only. However as they are low cost and have other benefits noone can really argue with them.

Feeder buses and park and ride are more controversial. People see that station car parks fill early in the morning peak so consider that more parking spots equals more patronage, so is a Good Thing. They might also look at the loadings on nearby buses and conclude that in a car-owning society few people will catch a bus.

Users of park & ride seldom pay for this privilege. Opposition politicians go crying to the press if even the possibility of charging is raised in some departmental briefing. However free Park & Ride is effectively giving one groups of people use of some very expensive land for 50 hours a week. This might not sound so bad until it is realised that by locking away this land we are denying other people more profitable and socially useful uses for it.

Such alternative uses could include additional retail, which provides services and creates local jobs. Or residential, which puts more people near a station and retail centre, so increasing usage of both. The more land for parking, the is less available for these 'higher and better' uses for the land.

Hence Park and Ride, particularly in established high-value suburbs, has a high opportunity cost that is not always realised by its proponents. Before it gets the nod, it should be compared to alternatives (notably better buses and better uses for the land) since these are likely to bring about better outcomes.

Now even if some scrap of land is deemed unsuitable for an non-parking purpose, commuter Park & Ride is not necessarily a no-brainer for that either. Local retailers generally want more parking, but hate long-term commuter parking.

The reason? A park & ride passenger is going to be 9 - 10 hours away working. In contrast a short-term 2-3 hour parking spot is going to be used by shop customers throughout the day. Hence there's more people and more business spread over the day, which must benefit local business more than an absentee parker. Though there are issues with local traffic management, short-term parking is clearly financially better for local centres than long-term commuter parking.

Then there is the scale that park & ride requires before it can substantially boost suburban rail patronage. Park & Ride requires heaps of land to benefit comparatively few people. Most people who catch the train in Melbourne continue to walk to their station. Enlarging P&R might encourage more beyond walking distance to catch the train, but the increase isn't going to be that great. That is unless there is a huge (and expensive) increase in P&R space, which in established high land-value suburbs will usually be at the cost of more productive uses and urban amenity.

Improved buses have none of these disadvantages. A bus carrying 30-50 people arriving every 10 minutes makes far better use of its bus bay than the handful of people who'd otherwise be parking there.

An improved bus service isn't cost-free either, but when compared to the increased value and trade from the more productive use of land, it looks a lot more attractive. And there's some elements of improved buses that are extremely cheap that we haven't done that well in Melbourne to date.

An example is bus service information at stations. A bus timetable installed at a station might cost between $1 and $1000 (depending on if a timetable case is needed) yet may benefit 100 alighting passengers. In contrast, one extra parking spot might cost $8000 (similar to a secondhand car) and benefit just one passenger per day. These benefit ratios of ten to one thousand to one are too large to ignore and tend to favour improved buses.

Timetable co-ordination to slash waiting and overall journey times is another low-cost measure. Wider spans and frequency cost a bit more, but again if opportunity costs of P&R are considered then it might even work out cheaper, as well as bringing other benefits such as better access to the local centre, urban amenity and a better-used bus service with reduced per-passenger costs and carbon emissions.

An objection to better buses is that most people have a car and most will use park & ride. This may be true with current bus services and levels of co-ordination. However it it not inherently true with every transport system, even in affluent high-car ownership cities.

Perth is an example. It has higher car ownership than Melbourne and purpose-built park & ride facilities at railway stations. Despite this, the proportion of train passengers who arrived at their stations by bus is far higher there than here. While Perth's relatively smaller rail network may play a part, the main reason is that Perth has made some effort with passenger information and service co-ordination. There is nothing to stop us from introducing similar measures, and I believe good along these lines will come from the current bus reviews.

To summarise: Park & ride does have benefits for some passengers. Some passengers will always use P&R even if the buses are good. However big expansions of it in built-up high land value areas are not without considerable penalty, opportunity cost and foregone revenue that its proponents tend to ignore.

Bigger bus improvements will naturally cost more relative to smaller bus improvements. However they also offer other benefits since they do not contribute to the traffic and land use issues that reduce the attractiveness of P&R.

While a balanced site assessment of the pros and cons of each needs to be made, I believe that in most cases, at least for built-up areas, a 'better buses' option is likely to produce greater overall economic, social and environmental benefits than a purely park and ride approach.