Cars & traffic

Jul 20, 2016

Should commuters pay to park at the station?

Commuter parking at outer suburban railway stations is usually free, leading to peak period shortages. The first step should be to charge for it

Alan Davies — Editor of The Urbanist

Alan Davies

Editor of The Urbanist

Leppington railway station, south west Sydney

It only opened in February 2015, but parking is proving to be a problem at Sydney’s Leppington railway station on the new South West Rail Link. The Sydney Morning Herald reports that the station’s 850 parking spots “are often full by 10am, leaving motorists who arrive later to circle repeatedly to find a park”.

This is a familiar problem at outer suburban railway stations across Australia like Melbourne’s South Morang station which opened in 2012. Peak-hour commuters elect to drive to them rather than take feeder buses. The choice is facilitated by under-priced parking – it’s usually free – and by uncompetitive feeder bus services.

What to do? Here are four possibilities.

Provide more parking

This approach recognises that driving is a very attractive choice for commuters in outer suburbs where traffic congestion is relatively low. While governments are reluctant to charge, it’s usually reasonably cheap to expand at-grade parking at outer suburban stations – circa $10,000 per space – because there’s often little development nearby.

Demand for commuter parking will increase over time but it isn’t infinite because only a small proportion of outer suburban workers have jobs that can be accessed quicker by rail than by road e.g. jobs in the city centre.

The cost of existing and additional spaces could be fully recovered by charging. However although parking meters are ubiquitous in Australian cities, governments and commuters seem to think the virtue of taking the train provides moral justification for driving to the station and parking at taxpayers’ expense.

Improve feeder bus services

Like providing free parking, this approach is costly too because the scope for cost recovery is limited. The catchments of outer suburban stations are often radial, requiring an effective feeder bus service to have multiple routes operating at high frequencies e.g. coordinated with train departures in the AM and arrivals in the PM.

Moreover, the potential market is small. For example, there were a total of 1,562 boardings in the weekday morning peak in 2013-14 at Melbourne’s outer suburban South Morang station. The relatively small pool of bus users increases the cost per passenger and reduces the environmental benefit per passenger.

Buses generally aren’t attractive to outer suburban commuters who already have a car because in most cases they’re slower and less convenient. Only those who don’t secure one of the existing stock of free parking spaces – or who can’t kiss-and-ride – are likely to be candidates for using feeder buses.

Provide safe bicycle routes

Bicycles and scooters offer many of the benefits of driving, especially with power assistance. Their weakness though is exposure to weather and, especially, concerns about safety. There’s no scope for directly recovering the cost of building a feeder network of safe (mostly segregated) routes, but the ongoing costs to both taxpayers and travellers would be very low.

Promote vehicle-sharing

It’s tentative at this stage, but there might be more scope to apply the idea of car-pooling to short feeder services – which have a common destination – than to the full commute, perhaps based on something like the Uber model.


What’s missing in the evaluation of these sorts of policy choices is reliable data on the relative financial and economic costs of the various options to both travellers and government. What can be said though is that a large part of the financial cost of using cars for feeder transport is carried privately i.e. by owners. The exception is under-priced government-provided parking.

Most of the cost of feeder buses, on the other hand, is borne by taxpayers since fares cover only a small proportion of capital and operating costs. A key consideration though is that governments must provide some level of public transport anyway to meet the needs of non-motorists.

The usual advantage of public transport over cars – i.e. externalities – isn’t large in this situation and might even be negative. Congestion is commonly the largest negative externality but it’s usually low around outer suburban stations. Moreover, the sorts of frequencies required to operate an effective feeder bus service, combined with the relatively low potential pool of bus users, means the environmental footprint per passenger is likely to be large.

The first step must be to charge for commuter parking at railway stations. That should recover the full cost of providing parking (up to circa $45,000 per space for multi-story parking); lower the demand for driving; and make the other options relatively more attractive. I suspect charging for at-grade parking is actually more politically plausible than timid politicians are prepared to allow (Transperth charges $2 per day).

Won’t charging for parking discourage use of rail for the journey to work? Yes, but I expect the shift wouldn’t be large. Most outer suburban workers who use the train – and they’re only a small proportion of all commuters living in the outer suburbs – do so because it offers a very big advantage over driving all the way.


Leave a comment

Your email address will not be published. Required fields are marked *

14 thoughts on “Should commuters pay to park at the station?

  1. Roberto

    Sydney bike commuters have had to pay for bike lockers and now for a space in new enclosures, by Opal card. This is at least convenient, but it seems inequitable that car drivers don’t have to pay. In the Inner West we have unrestricted parking on local roads near light rail, when the whole point was to have tram users walk or cycle to the stations. This parking often precludes installation of bike paths to the station.

  2. Alan Davies

    Tony Morton

    I don’t think you can buy and operate 3 buses and nine drivers for a mere $2,800 per day. Even at a conservative operating cost per bus of $125 per revenue hour, the total daily cost over 18 hours comes to $6,750. Then there’s the capital cost of each bus…$300,000 to $500,000.

    Capturing an average of 26 passengers on each of the 108 daily services you’re proposing is optimistic to say the least when the average passenger load per bus in Melbourne at present is circa seven. There’s no hope off-peak (evenings and weekends) because there’s plenty of spare, free car spaces at those times.

    The $1 cost you ascribe to bus passengers is notional; it’s not an increase in revenue. If you ascribe it to the bus leg then you have to deduct it from the rail leg. Revenue from parking on the other hand is real money. What matters for the financing of the parking lot is the borrowing rate; governments don’t pay anything like 7%.

    PS. How many 5km x 0.8 km bus services will a new station like Mernda require? Will one be enough by 2020? 2025?

    1. Tony Morton

      Alan, I’m sure you don’t mean to say bus drivers get paid anywhere like $100 per hour, but you’re making it sound that way. The figures that matter here are the marginal operating costs, not (as in the source cited) the total costs including agency and government head-office operations, because the latter bear no relation to how many bus route-kilometres we operate.

      I’m basing costs on an ATRF paper by Graham Currie, Ray Kinnear and Ian Wallis, adjusted for inflation. A bus costs around $400,000 which is equivalent to $35k per year in cash flow terms if it lasts 20 years: I’ve added another $35k per year in depot/O&M costs. Route operations are $40-50 per hour of which around $10 is fuel and the rest is driver wages and on-costs. (FWIW, a quick Google search reveals that the average Australian bus driver wage is $24 per hour.)

      I attributed $1 of the Myki money fare (of between $1.35 and $3.90) to the bus service precisely because many trips have both a bus and a rail component (with the higher fares much more likely to have a longer rail component). But again, if feeder buses are delivering at least some passengers who wouldn’t use public transport if the bus service didn’t exist, then attributing more of the additional revenue to the bus service is warranted. Especially since the marginal cost of conveying an additional rail passenger outside peak hour in the peak direction is virtually zero.

      The ‘optimistic’ patronage capture of 2,800 person-trips per day across 108 services is based on a low-ball population density and a mildly ambitious mode share target (22% of a 6,400 catchment population out and back each day). What it really assumes is that the service is designed and promoted as something everyone can use for whatever purpose any time, much as it is in Europe – rather than at present where we run something we know to be uncompetitive with car travel and regard actual passengers as an optional extra!

      Note too that the 7% is a discount rate, not an interest rate. It accounts for the fact that when you run a feeder bus, the operating costs fall due in future years, whereas a car park requires almost the entire expenditure to be made here and now.

  3. Dylan Nicholson

    It’s hard not to be a bit concerned that a portion of travellers who currently park & ride might simply switch to driving the whole way based on the total cost. There’s not much point having a disincentive to park at a station unless they have better alternative options, whether it be cycling/walking there, being dropped off or using some sort of feeder service.
    I certainly think a lot more could be done to encourage people to get to train stations/major bus exchanges by bicycle, obviously including decent and secure parking.

  4. Roger Clifton

    Of course we should pay. But please make it credit card, as coins drive me crazy and cost money to collect.

    Sudsidizing parking undermines the profitability of taxis lurking there and the likelihood of setting up minbus loops through the suburb around each station.

  5. Peter

    Perth started with charges for new carparks at new stations. The Perth SmartRider card could be used to pay the parking charge. In 2014, paid parking was extended to the entire TransPerth rail network. The flat all day fee is $2.00

    New car parking can easily cost in excess of $40 000 per car parking space, there should be a charge for it because without a charge, the carpark will fill up.

    See Free rider over for Perth Commuters

  6. Daniel Bowen

    “Moreover, the potential market is small. For example, there were a total of 1,562 boardings in the weekday morning peak” – that’s the current figure, but it’s constrained by the limitations of park+ride (despite an above-average sized car park) and inadequacies of feeder buses and cycling/walking options. The question is, what is the potential if some of those limitations can be overcome?

    (In the case of South Morang, the rail extension to Mernda will obviously change the situation somewhat.)

    1. Alan Davies

      It won’t be zero, but I suspect it’s not large either; after four years, South Morang commuters have worked out whether or not taking the train to work makes more sense than driving. Some start earlier in order to get a park while others, like Amark, will fall back on kiss and ride. Haven’t got the numbers to hand, but kiss and ride accounts for a surprisingly large proportion of outer suburban commutes.

      1. Heavy Rail

        As a rule of thumb, we often assume 40% of car access trips are KnR in the outer suburbs, and 60% of car access trips are KnR in the inner suburbs.

        In the context of Rail station entries, 1,562 in the morning peak 2.5 hours is quite large, it puts South Morang in the busiest 20% of metropolitan rail stations. Around 60% access by car (lets say just under 1,000), and there are only 450 off-street car parks, so that implies that there are quite a few KnR trips, and a lot of money to be made by charging for parking.

      2. Tony Morton

        Put it another way: there are 1,562 AM peak boardings and 3,250 total weekday boardings at South Morang station at the last tally. In 2010 there were zero boardings – did this mean the market was nonexistent? It was argued in the past that extending the Epping line would only cannibalise patronage at existing stations (much as it’s still argued a Doncaster train line would get almost all its patronage from the existing network). Yet the worst case displaced weekday patronage from the year-on-year figures is 1000 from Epping, 500 from Thomastown, and 100 each from Keon Park and Lalor, accounting for at most half the current boardings at South Morang.

        The point is that simply by extending good public transport service closer to residents on the urban fringe, new patronage and revenue has been generated. It’s not outlandish to expect a similar effect from better-quality feeder bus services that extend the walkable catchment within reach of good public transport. Currently, the rail network actually gets the majority of its patronage from people living within walking distance of stations, because supporting bus services are inadequate and car parking has too many inherent limitations. We can contrast it with Toronto where three-quarters of suburban subway travellers access the station by connecting bus or tram.

        The other thing to remember about feeder buses is that they can perform a dual role in conveying people to local district centres (which fortunately in Melbourne are often co-located with railway stations).

        1. Alan Davies

          The comparison with the proposed Doncaster rail line is misconceived. Rail patronage increased on the Epping/South Morang line because the population of the region is growing rapidly; it’s a Growth Area.

          However there’s very little population growth in the Doncaster region so there’s no low hanging fruit. Even with some consolidation, the population of the City of Manningham grew by just 1.3% over the five years from 2006-11; that’s lower than the modest 2.6% growth it recorded over 2001-06. It’s projected to grow between 2014 and 2036 at an average of just 0.77% per annum.

          As I noted above, the key thing we’re missing is data. The back of an envelope cost to government of providing 500 new at-grade parking spaces is $5 million, which could be recouped over time with a modest parking charge. What would it cost to set up and operate a feeder bus service that is (a) dense enough to put 95% of households within 400 metres of a bus stop, (b) coordinated with all-day train departures/arrivals at (say) South Morang, and (c) attractive enough to win travellers away from driving?

          1. Tony Morton

            Indeed it’s a growth area, but as the residents will tell you, most of the growth around the South Morang station catchment actually occurred a decade ago. Funnily enough, a decade ago is also when Plenty Road had all the work done to widen it, but the residents had to wait years longer for rail service.

            The post-2010 growth is mainly in Mernda further out. As things stand, ‘demand’ for public transport in Mernda is suppressed by the fact one has to travel so far to access a frequent service, and many people are already halfway to their destination by the time they get there. The new improved bus service from next week will help, but even that only provides service every 20-40 minutes outside peak hour.

            I did a back of the envelope calculation of feeder bus costs in a presentation to the Victorian Transport Infrastructure Conference a month or so ago. Basically, if you take the example of a feeder route with a 10km round trip, taking 30 minutes at typical bus speeds, you can operate a 10 minute frequency with 3 buses and 3 driver shifts per bus over an 18 hour day. The total cost works out as just on $1 million per year, and breaks down as about 20% capital and 80% operating expenditure. The daily cost is around $2800, which would be fully recovered with an average of 26 passengers per run, assuming $1 Myki money revenue per passenger is attributable to the bus service. If the catchment area measures 5km x 800m with a population density of 16 per hectare (for 6400 people total), you can achieve this patronage level if 22% of the population make one trip out and back each day, and the other 78% don’t use public transport at all.

            Now, that bus route can deliver people to the station and local district centre all day, but if we just take the number of passengers delivered between 6am and 8am at the break-even patronage level – 312 in total – and build them a car park instead, the minimum cost is $3.1 million (the same as operating the bus route for 3 years). To recover the cost at a 7% discount rate with a 20-year economic lifetime, you would need to charge about $880 per space per year or about $3.60 per weekday in addition to the train fare.

            And the catch with the latter is that as soon as you want to expand it, the cost will escalate to more like the $43,000 per space we’ve just seen at Syndal station, which would require charging $16 per weekday just to break even – and for which the cost of a 312 space car park exceeds the entire present value of that 10km feeder bus route.

            It all comes down to whether we’re providing public transport with the aim of attracting a significant proportion of the population as customers, or merely as a supplement to car dependence, with built-in limitations that waste most of the capacity of the infrastructure.

  7. Amark

    parking spots “are often full by 10am”! what a luxury! My station car park is always full by 8am at the latest. I’ve gone back to being a kiss and ride person

    1. phonakins

      not on the central coast by any chance?

Share this article with a friend

Just fill out the fields below and we'll send your friend a link to this article along with a message from you.

Your details

Your friend's details