David Gargett

1
PASSENGER MOVEMENTS BETWEEN AUSTRALIAN
CITIES,197071 TO 203031
?

• David Gargett
• Afzal Hossain
• 13 February 2007

2
Background

• Arose from need to forecast light
• vehicle traffic on the national highways.
• But need to consider all modes.
• Thus aim is to model drivers of total
• passenger travel between cities.
• Only then can one predict highway car
• travel patterns.

3
Objectives

• Estimation of passenger movements by mode between
  Australian major cities from 1970-71 to 2003-04.
• What will growth be to 2030-31 in passenger
  movements by mode between Australian cities?

4
Ten city pairs

• Sydney-Melbourne
• Sydney-Brisbane
• Melbourne-Brisbane
• Melbourne-Adelaide
• Eastern States-Perth
• Sydney-Adelaide
• Melbourne and Sydney-Gold Coast
• Eastern States-Tasmania
• Canberra-Sydney
• Eastern States-Northern Territory

5
Modes

• Air
• Car
• Coach
• Rail
• Other

6
Time series

• Historical Data 1970-71 to 2003-04
• Projections 2004-05 to 2030-31
• Based on
• GDP (Treasury)
• Population Growth Rates (ABS)
• Changes in Fares (various sources)

7
Main data sources

• Designed around tourism data on inter-regional
  passenger movements.
• Because the data has been and continues to be
  measured by Tourism Research Australia (TRA), to
  the tune of 4M/year.
• TRA surveys both domestic and international
  travellers about their travel between the regions
  of Australia.
• This data has been assembled for the 10 city
  pairs by mode from 1970-71 to 2003-04.

8
Gravity Model

• Total passenger travel between any two cities
  (say i and j) can be calculated
• Tij (Pi x Pj x GDPc2)0.524 / (Tc / CPI)-0.565
• where
• Tij - Total trips between regions i and j.
• Pi and Pj - Total population in region i and
  region j.
• GDPc - National gross domestic product per
  capita.
• Tc - Real generalised cost of travel.
• CPI - Consumer Price Index.

9
Logistic Substitution Model

• Forecasts of total travel were converted to
  forecasts for specific modes by using logistic
  substitution models of mode split.
• For each mode, a competitiveness index was
  estimated based on changes in the mode share over
  the last decade.
• A competitiveness index below 1.0 means the mode
  is expected to decline over time in share
  relative to air.

10
Logistic Substitution Model Contd.

• For example, on the short Canberra-Sydney route,
  the competitiveness indices are
•  
• Air - 1.00
• Car - 1.02
• Coach - 0.98
• Rail - 1.03
•  
• On the long Sydney-Brisbane route (more typical
  of the intercity routes), the competitiveness
  indices are
•  
• Air 1.00
• Car - 0.97
• Coach- 0.93
• Rail - 0.93

11
Passenger Movements by Mode
12
Mode Share (per cent)
13
Average annual growth rates by mode,1970-71 to
2003-04
14
Average annual growth rates comparison (all
modes)
15
So what do we get ?

• An understanding of the drivers of travel.
• An understanding of the patterns of mode share
  change.
• A link to a continuing and funded data source
  updated yearly.
• 1 to 3 provide the basis for the OZPASS
  interregional travel model.
• Forecasts out of the OZPASS model are providing
  forecasts of car traffic along AUSLINK Corridors.

16

• These forecasts are
• Systematic based on transparent research.
• Validated Bruce Highways over 10 years.
• Multi-modal e.g. current airports project.
• Open to scenarios i.e. policy changes or
  changes in assumed conditions (e.g. fares).

17
Summary

• On all routes, except CanberraSydney route, air
  travel has been progressively taking mode share
  from car plus coach and rail.
• This effect will be less important in the future
  as the rate of mode share capture by air slows.
• Overall, total passenger travel growth is
  expected to continue to grow more quickly than
  GDP.
• The relationships found on the 10 corridors have
  been built into the OZPASS interregional travel
  model.

18
FOR A COPY OF THE PUBLICATION VISIT

• www.btre.gov.au/Publications/Information
  Sheet/
• Information Sheet 26

19
Thank you

• Any question ?