Home  About The Rail Now Campaign  Why we need Rail Now  Our Track Record  Links  Members 

You can download the following document in Adobe PDF format here.
(Internet Explorer users right-click and click "Save Target As". Netscape Navigator users right-click and click "Save Link As".)

Thursday 15th October 2004

Article

Paying the Rent

If Sydney's land is so valuable, and rail uses land with greater efficiency than roads, why does Sydney rely so much on roads?

With so many demands on Government funds, it's important to get value for money from our transport systems.

In economic terms, value for money means rate of return. Do we get a good return for the resources invested in our roads and railways?

One of these resources is publicly owned land. Sydney land, as any home owner knows, is very valuable(1). Since land is so valuable, we should ensure:

  • That we waste as little of it as possible on transport corridors; and
  • That we use our transport corridors as efficiently as possible.

Is this what our Governments do? No.

It is difficult to find an accurate estimate of the area covered by roads in Sydney. But it is plain that the area covered by roads is substantial, and that covered by railways is much less. In fact, a study by the Institute for Sustainable Futures in 1999 concluded that the value of land under roads in Sydney was 22 times higher than the value of land under railways(2).

Yet railways have a far greater carrying capacity than roads. That is rail's great economic advantage. Rail delivers far more passengers, or tonnes of freight, per square metre used than any other mode of transport.(3)

So we have a paradox: if Sydney's land is so valuable, and rail uses land with much more efficiency than roads, why does Sydney rely so much on roads?

The reason is that our economic assessment processes are not nearly as rational as we would like to believe. While several examples from individual project assessments(4) could be presented, this article considers the overall picture.

One reason Sydney relies on roads is that users do not pay to use transport corridors.

  • With railways, for the last few years State Rail has paid an Access Charge to another Government entity for its use of the infrastructure. Theoretically this charge could have included an amount for the rent of the land, but in practice it hasn't, due to an IPART recommendation in 1999 that the rail corridors be valued at zero for the purposes of the NSW Rail Access Regime.(5)
  • With roads, petrol taxes and registration fees have never been levied by reference to the cost of providing roads.

Accordingly there is no connection at all between the value of our transport corridors, and the price we pay to use them. Users do not pay for the use of transport corridors. Use is subsidised by the taxpayer.

Since Governments are so concerned about transport subsidies, why doesn't the Government require a rate of return on the land the community has invested in those corridors? If the Government used taxpayers' money to acquire an office block, and then allowed people to use the offices without paying rent, it would rightly be condemned as wasting an asset. But this is exactly what we do with transport corridors. Is that sensible management of a community resource worth billions?

Not only is there no method of collecting the rent on transport corridors, there is no official recognition that this is necessary or desirable. Those who determine our transport policy choose to ignore the issue.

  • Subsidies for train and bus travel are a constant theme of State Government announcements, but when has the Government ever mentioned the rent foregone on transport corridors? Why isn't that a subsidy?
  • Local Governments pay for the upkeep of local roads, but they do not even keep a record of the roads' value in their accounts(6), and have no power to charge for their use.(7)
  • IPART has for years focussed on the level of cost recovery from commuters, while ignoring the level of cost recovery from motorists. Rail's superior rate of return on land can therefore be ignored. IPART has never specified a criterion by which to assess the appropriateness of any subsidy to rail(8). IPART's chairman, Professor Parry, recently completed an inquiry into transport for the New South Wales Government(9). Nowhere does it mention the rate of return on land invested in transport corridors as an issue.
  • Studies on whether motorists are subsidising the Government through petrol taxes, or vice-versa, sometimes omit as a cost to Government the rent foregone on the community's land. Studies which include it assess subsidies on a national basis, whereas the superior rate of return on land of railways is most evident in our crowded cities, where land values are highest. Subsidies to different transport modes should be measured at the level at which they affect travel behaviour - the regional level.

The prevailing orthodoxy is to regard the land in existing corridors as a 'sunk cost' which can be disregarded. A minority view is that the use of land for transport corridors imposes a cost on the community in the form of the foregone opportunity to use the land for another, more profitable purpose(10). This argument assumes that public land is available for members of the public as a transport corridor, so the only real cost is the 'opportunity cost' involved in not putting the land to a better use. As a cost imposed on others, it is regarded as an externality.

What you measure, and how you measure it, depends on the purpose of the measurement. It is hard to see any purpose which would justify treating valuable land as an irrelevant 'sunk cost'. Where the purpose is to examine the relationship of the transport sector to the economy as a whole; for example, when the question is whether public land should remain as a road or become a park or a hospital site; then the 'opportunity cost' approach to land may be justifiable.

But when the purpose is to better manage the transport system, both the 'sunk cost' and 'opportunity cost' approaches ignore the obvious - you cannot provide transport without land, so the cost of making that land available is a direct cost of providing the service. It is a direct cost incurred by the producer which is not, at present, passed on to the consumer.

The failure to collect the rent on the transport corridors has these consequences:

1. The supplier of transport corridors, the Government, has no effective mechanism for rationing the private use of public space.

2. The Government has no economic incentive to maximise its rate of return on the land invested in transport corridors, by switching to the optimum mode for each route.

3. Users have no economic incentive to make efficient use of roads.

All this favours the mode which is the least efficient user of land - namely travel by private car. As a result, Sydney has more roads than is economically optimal, or ecologically sustainable.

We are not getting value for money on the land we've invested in roads and railways, because we don't count all the dollars being spent on transport. We simply forget about collecting the rent on the transport corridors.

If this city wants a rational, sustainable transport system, the users of transport corridors must pay the rent for their use of the corridor. How the rent would be collected will be dealt with in subsequent articles.

Written by Philip Howell for the Rail Now Campaign Inc.
howell@bigpond.net.au

Footnotes
(1) The value of land subject to council rates in Sydney was $466.5 billion as at 1st July 2003. Note that this is the unimproved land value which excludes the value of houses and other buildings, and excludes the value of transport corridors. This figure is derived from data supplied by the Valuer-General of NSW. See Appendix 1 to "Funding Sydney's Transport - A Sydney Transport Levy", by Philip Howell, 5th November 2003 at www.railnow.org.au. Since GDP for the year ended 30th June 2003 was $731.022 billion (see ABS, National Income, Expenditure and Product, 5206.0, June 2003) Sydney land values at this time were worth more than 60% of GDP for the year ending on 30th June 2003.
(2) The study is "Sydney's Passenger Transport: Accounting for Different Modes", by Banfield, Hutabarat & Diesendorf, ATRF99, Papers of the 23rd Australasian Transport Research Forum, published by the Western Australian Dept of Transport. The figures used here come from working papers for the study. The value of land under all roads in Sydney in 1996 was $39,168 million (see p.10) and the value of land under railways was $1,765 million (p.15). These figures were derived partly from road length and value figures published by the Roads and Traffic Authority and track length and value figures published by the State Rail Authority. Since both are State Government entities, it is likely their land valuation methods were the same, which implies that the area of land under roads in Sydney is 22 times greater than that under railways. Key assumptions in the study's calculations included:
# that the length and value of RTA roads in Sydney was proportionate to the ratio of Sydney's population and vehicle fleet to that for New South Wales as a whole (p.9),
# that local roads were on average 2/3 as wide as classified roads (p.10), and
# that the use of land by Sydney's CityRail system was in proportion to CityRail passengers' use of the SRA's whole network (p.15).
(3) For passenger travel, railways are often planned on the assumption that each 8 car train can carry between 800 and 850 people. With 15 trains an hour being achievable on many Sydney railways, one railway track can move more than 12,000 people per hour. The Christie Report in June 2001 stated that average train capacity should be assumed to be 1,000 passengers: see p.29; meaning 15,000 can be moved each hour by one railway track. Compare this with motorways. In the table below, the two columns on the left show traffic volume data from the Roads and Traffic Authority for selected motorways in 2002. The second column records the average weekday traffic heading in one direction for the busiest hour of the day. The data does not reveal the number of passengers moved by this traffic. This is estimated (by the author) in the two columns on the right, using an assumption that there will be between 1.12 and 1.2 people per car in the peak period. These are the figures the RTA itself uses when planning roads - see Roads and Traffic Authority Economic Analysis Manual, Version 2, 1999, appendix B, Economic Parameters for 2003, Table 8.

Road (Location)No. of VehiclesNo. of People At 1.12 per VehicleNo. of People At 1.2 per Vehicle
M4 Homebush (Station 28.001.E)3,6404,0764,368
M5 Moorebank (Station 60.002E)4,9585,5535,949
M2 East Ryde (Station 52.039.E)2,5142,8163,017

While these figures are indicative only, and of course omit freight, note that the motorways referred to each have 2 or 3 lanes, yet move substantially fewer people per hour than a single railway track.
(4) Some examples are presented in a paper available from the author, "Economically Sustainable Transport", by Philip Howell, 20th October 2001.
(5) Recommendation 8 on p.30 of "Aspects of the NSW Rail Access Regime", Independent Pricing and Regulatory Tribunal, April 1999.
(6) The relevant accounting standard, AAS 27, states in paragraph 108 that local governments "may elect ... not to recognise land under roads as an asset in their statement of financial position" until the end of 2006.
(7) Section 5(1) of the Roads Act 1993 (NSW) gives the public rights to pass along public roads subject to such restrictions as are set out in legislation. Since no legislation authorises councils imposing a charge for the use of roads, they could not do so without infringing the right given in section 5(1). Under the Local Government Act 1993 (NSW), councils can raise money in any of the following ways:

  • By levying ordinary or special rates, and domestic waste management charges, but only on "rateable land" - ss.494, 495. Land vested in a Crown authority, such as roads owned by the Roads and Traffic Authority, is not "rateable land": s.555(1)(a).
  • By levying charges for services provided on an annual basis - s.501(1) - but the provision of roads is not a "service" for which such a charge may be made: s.501(1) and clause 8 of the Local Government (Rates and Charges) Regulation 1999.
  • By levying fees for services provided on a non-annual basis - s.608 - but roads are provided on an annual basis.
  • By charging fees for certain other activities, but the provision and maintenance of roads is not one of these activities: s.610A(1), s.611(1).

(8) IPART increased fares for years on the basis that rail users were being subsidised too much, without ever identifying a criterion by which an appropriate level of subsidy could be determined. It ignored the author's submission dated 14th April 2000 that the logical starting point for setting the level of cost recovery from rail, is the level of cost recovery from roads. If the two are not equated, the allocation of resources will be distorted in favour of the more heavily subsidised mode.
(9) Professor Tom Parry conducted the Ministerial Inquiry into Sustainable Transport in New South Wales. Chapter 8 of his final report, dated December 2003, entitled "Charging for Road Use", (pp.72-82) acknowledged that road users do not pay to use roads, and that this disadvantaged public transport. However, at no stage did the Professor acknowledge the rent, or rate of return, foregone on the transport corridor as one of the costs. His whole discussion is about pricing roads so that users pay the full cost of externalities - congestion, accident costs and pollution. For Professor Parry, the land is a 'sunk cost' which can be ignored.
(10) See the discussion paper commissioned by the NSW Independent Pricing and Regulatory Tribunal, "Subsidies and the Social Costs and Benefits of Public Transport", by the Centre for International Economics, March 2001, pp.18-21, available at www.ipart.nsw.gov.au.