The Kirby Reports: Volume II - Report (A. Transport Criteria

-117-

VI FORECASTING

1. FORECASTING FUTURE TRAVEL

1.1 Why Are Forecasts Necessary?

In a Report to the Minister in February 1976, the
Urban Transport Advisory Committee (URTAC) ₍₆₂₎
suggested the following as a principle to guide
the Department of Main Roads and the Planning and
Environment Commission ₍₆₃₎:

"Concentration on existing problems of
immediate rather than long term benefits."

In concentrating on the present, rather than future
traffic problems, New South Wales has adopted a
strategy also adopted by the United Kingdom ₍₆₄₎
and the United States ₍₆₅₎:

The construction of a major road is a vast project.
The estimated time for construction of the Cooks
River Route is 7 years. The same estimate is given
for the South Western Route. They each may well
take more. Even though the road building authority
has been enjoined to concentrate on existing
problems, it is not suggested anywhere that the
future can be neglected. The Transport and Town
Planners must obviously keep their eye on the
future, and make provision for major roads where
the need is demonstrated

There is no escape, therefore, from the need to
forecast the future. The Leitch Committee makes
the point in this way (although its words may be
more appropriate for a rural road than a major urban
road) ₍₆₆₎:

62. Now known as TRANSAC.
63. Submission S.K/C 340 DMR July, 1979, page 3.
64. White Paper 'Policy on Roads' 1978, page 15.
65. Car Cult Country K.W. Dobinson, page 17.
66. Report of the Advisory Committee on Trunk Road
Assessment (The Leitch Report) October 1971, page 8.

-118-

“Trunk roads have long lead times.
Typically..7-15 years may elapse
between inception and opening and
this period has been increasing in
recent years. Trunk roads are also
long lived assets and must be able
to cater for traffic levels well into
the future. The Department has adopted
the convention that a scheme should
have sufficient capacity to cater for
traffic levels 15 years after its
opening. Added to the lead time,
this means that forecasts must be
made of traffic levels 25 years ahead.
Thus schemes currently in their initial
planning stages require forecasts of
traffic beyond the turn of the century."

1.2 The Hazards of Forecasting

Forecasting is a hazardous business. The forecaster,
like the juggler, is required to keep so many balls
in the air that almost inevitably one falls to the
ground. Sometimes the entire forecast falls with it.

The forecast will, faute de mieux, rely upon trends.
Trends have a habit of changing. The symptoms of
change may not be evident until some time after the
change has taken place.

A number of illustrations can be given. We have
referred already ₍₆₇₎ to the failure of the
Cumberland Plan to foresee the staggering post-war
population growth in Australia. That failure was
not due to any lack of foresight. It was essentially
unforeseeable. There was a change in direction
brought about by a change in policy.

The Sydney Region Outline Plan had the misfortune
to make its prediction (in 1968) at a time when the
population curve had suddenly (and imperceptibly)
altered course. The result was a substantial over-
estimate of the likely Sydney population in the
2000.

67. Page 45.

-119-

The Sydney Area Transportation Study had the added
misfortune to commence its task before the signs of
the reduced population growth had properly surfaced.
It adopted the estimates of the Sydney Region
Outline Plan making some minor adjustments to give
expression to a policy of decentralisation ₍₆₈₎.

It is somewhat unkind of this Inquiry to remind the
SATS Inquiry of the prediction the latter made
concerning its own recommendations. It said ₍₆₉₎:

"The range of data collected and the
scope of the Sydney Area Transportation
Study recommendations are impressive
and of such importance to Sydney that
it is difficult to see this Report ever
being put under wraps. In particular,
the basic corridor proposal could give
Sydney a flexible and stable long-term
transportation plan."

The hazards of forecasting are such that even before
the publication of the Sydney Area Transportation
Study it was overtaken by events. The State Planning
Authority published (1973) a revised population esti-
mate. It substantially departed from that previously
given by the Sydney Region Outline Plan upon which
SATS had relied ₍₇₀₎. Within moments of that publica-
tion, there was yet another publication. It further
revised downwards the population forecasts ₍₇₁₎. The
SATS recommendations are well and truly under wraps.

Within a year of the publication of the Sydney Area
Transportation Study the Urban Transport Study Group
(now STSG) was asked to suggest modifications to the
recommended network. It was asked to reflect the
more recent information on population _(71A).

68. SATS Volume 2, ChaPter III, page 1.
69. SATS Volume 2, Chapter VIII, page 10.
70. SPA Technical Bulletin No. 4.
71. Addendum to Technical Bulletin No. 4.
71A. UTSG (now STSG) Revision of the SATS-Year 2000-
Travel Forecasts July 1975.

-120-

Since 1975 the position has changed further. The
population forecasts have been revised yet again.
Traffic predictions used in this Inquiry were based
upon population projections in a document published
by the New South Wales Planning and Environment
Commission in August 1976 (72). The most recent
figures effect a further downward revision (February
1979) ₍₇₃₎.

Even if the trend remains constant, the community’s
perception of the problem may change. This was
evident in the early 1970s with a so-called
'freeway revolt’. The number of motor vehicles
continued to increase at an extraordinary rate.
Yet, suddenly, people the world-around objected
(or appeared to object) to the freeway as a means
of accommodating traffic growth. There has been a
renewed interest in the environment and its preser-
vation. The public response to this Inquiry would
suggest a renewed interest in public transport as
a feasible solution to many traffic problems.

1.3 The Danger of Traffic Forecasts

The hazards of traffic forecasting are not simply
that they may be wrong. That is certainly one
hazard. It is one which was not concealed by the
Department of Main Roads. The Department says in
its submission ₍₇₄₎:

"Travel forecasting is not an exact science.
Like weather forecasting economic forecast-
ting, or forecasting of the result of a
horse race,, it is fraught with uncertainty.
Nevertheless, planning for the future
cannot realistically take place without
some estimates of future conditions being
made. Every individual makes plans in some
measure, and those plans are based consciously
or unconsciously on that individual’s inter-
pretation of how conditions will change from

72. Technical Bulletin No. 8 Where the Low Forecasts
    were adopted.
73. Population Projection for New South Wales 1976-2001
    PEC & See Discussion Paper Sept. 1979 ‘Sydney Region
    Population Estimates for Local Government Areas
    1976-2001 No 79/11.
74. S.K/C 340 DMR Transport and Economic Analysis
    September 1979, page 2.

-121-

those existing at the present. Society
also plans for the future, and again
forecasts are required so that realis-
tic plans can be made.”

The second danger is that a forecast will dramatise
deficiencies in the road system and conceal important
questions of policy. If a corridor of movement is
already straining to carry 10,000 vehicles in the
2 hour morning peak, and 20,000 vehicles are predicted
for 1991, it is difficult not to panic. It is
difficult to restrain the reflex to provide more
road space. Yet policy considerations should enter
the equation at two separate points.

First, the forecast is a prediction based upon an
interpretation of sign posts on display in years
gone-by. It postulates, to a greater or lesser
degree, a continuation of past trends. There is the
danger that it will ignore the planner’s ability to
moderate past trends by policy initiatives.

Secondly, even where there will be insufficient
capacity no matter what initiatives are taken between
now and the forecast year, it is still necessary to
contemplate the land use implications of meeting the
traffic demand. Is it a good thing for the city's
overall structure and growth if the evident deficiency
between 10,000 vehicles and 20,000 vehicles (to take
previous example) is corrected? We venture to suggest
that if the deficiency were one within a radius of
5 kilometres of the central Business District, and the
direction of travel was towards the City, everyone
would now agree that it is better to shrug one’s
shoulders, rather than attempt to correct it.

The policy question (as to the land use implications)
must be confronted no matter what the direction of
travel. It may be rather more difficult to answer
where the direction of travel is not (as in the
example) towards the Central Business District. Will
it lead to the expansion of the urban fringe? Will

-122-

it prejudice a policy of regional self-containment?
Is it necessary for the purpose of providing
reasonable accessibility for trucks between major
industrial areas, or between important traffic
generators? Can the deficiency be answered by
public transport? Would a road compete with public
transport facilities? Is there a land use solution
in the offing?

Traffic forecasts, because they highlight dramatic
differences between supply and demand, may deflect
attention from these important policy issues. In
the presentation of material to this Inquiry the
discussion has been almost exclusively in terms of
congestion, and its eradication, rather than the land-
use implications of its eradication, and the land-use
implications of providing a facility. The discussion
if this be a fair description, is an example of the
second danger.

We have demonstrated already that the system will
continue to operate even were the Department of Main
Roads to refrain from building any further roads ₍₇₅₎.
The road builders have the opportunity, therefore, to
finely tune the system and pursue matters of policy
to a far greater degree than planners in other arenas.

2. TECHNIQUES OF FORECASTING

2.1 Extrapolation of Past Trends

One obvious way of predicting the future is to monitor
the past. In this way one can discern the rate at
which traffic is growing. The Department of Main
Roads (in common with all road builders) employs this
technique. Certain counting stations have been
established throughout the Sydney Metropolitan Area.
They monitor the number of vehicles each day. They
are permanent stations, and are established on the
main arteries. There are, in addition, checks made

75. See page 8.

-123-

on every major road every two years. Figures,
known as Annual Average Daily Traffic (A.A.D.T.’s)
counts are compiled and published.

Reproduced below is a graph extracted from the 1977
A.A.D.T. Volume. It represents the traffic growth
on the Princes Highway as it, crosses the Cooks River
at Tempe (where there is a permanent station). It
will be seen that a maximum and minimum growth rate
is calculated ₍₇₆₎:

FIGURE 6.

GROWTH – PRINCES HIGHWAY AT COOKS RIVER BRIDGE, TEMPE

76. Traffic Volumes & Supplementary Data 1977,
D.M.R., page 44.

-124-

2.2 The 'Four Step' Model

The Department of Main Roads with the assistance
of the State Transport Study Group) also employs a
computer modelling technique known as the 'four-
step model’. It simulates traffic flows for the
whole of Sydney

The technique is complex. It has certain advantages.
It also suffers from certain limitations. Its very
complexity is liable to 'bamboozle' the public
unless steps are taken to explain its features and
the assumptions which underly it. We shall endeavour
to briefly describe it in the following chapter.

-125-

VII TRAFFIC MODELLING

1. THE TECHNIQUE DESCRIBED

1.1 Introduction

In the course of this Chapter we will deal with
the following:

The way in which computers can be employed
to reproduce the flow of traffic through the
Metropolitan road network (the construction
of a traffic model)
We will briefly describe the traffic model
used in the present Inquiry (known as the
'four-step' model)
We will examine the limitations inherent
in the modelling procedure
The modelling procedure makes certain assump-
tions as to population, employment distribution,
car ownership, fuel availability etc. We will
briefly examine those assumptions
Finally we will examine the way in which the
four-step model was employed in the present
Inquiry

1.2 Traffic is Regular and Therefore Predictable

The task is daunting. It is to replicate with the
aid of a computer the myriad journeys made each day
by ordinary citizens, and by commerce.

Yet there is a regularity, and hence, a predictability,
about the journeys which we, as individuals, make
day-in day-out, week-in week-out. We have in the
past, and we will in the future, get up at a certain
hour each day. We will journey between our home and
our place of work. We may, from time to time, change
jobs. Almost certainly we will remain within the same

-126-

job classification. If we do not journey to this
factory, or that office, we will journey to some
other factory or office.

The transport modellers seize upon that regularity.
They suggest it can be monitored by means of a
survey. The survey can isolate the variables which
either generate trips or attract people to an area.

1.3 The Model Reproduces the Morning Peak

The morning peak period (between 7.15 a.m. and 9.14
a.m.) is the period of greatest regularity. Time is
at a premium. Few people have the energy or time to
deviate from the shortest path between their home and
place of work. The journey to work is compressed
within a shorter period. Consequently the system is
under greatest strain during this period.

The morning peak, in this respect, differs from the
evening peak. In the evening there is far greater
flexibility as to the time of travel. Some people go
directly home. Others do their shopping on the way
home. There are others who congregate with friends in
bars and coffee shops. The result of one survey (in
Canberra) is described in the following paragraph ₍₇₇₎:

"About 26% of workers made one (or more)
major stop on the work to home movement..
Within this group, single stops were
most common, comprising 20% of the total,
and multi-stops comprised only 6% of the
total..It can be seen (from a table) that
over one-half of all stops were for
shopping or personal business purposes,
with social recreational purposes accoun-
ting for 25%."

The Sydney Area Transportation Study was ambitious.
It was concerned to monitor shopping trips and trips
made for recreational purposes (as well as peak period

77. A Study of Travel Linkaqes: Implications for
Urban Transport Planning, N.R. Graham and K.W
Ogden, 5th ATRF Forum Papers, page 443.

-127-

travel). More often than not, shopping and recrea-
tional trips are made in the off-peak. They are far
less predictable.

They are affected by random factors such as the
weather, other business to be transacted and chores
to be performed by the person making the journey.
They also reflect that person’s assessment of the
most convenient time for the trip, to avoid congestion
and waiting.

The State Transport Study Group (which is responsible
for traffic modelling) has chosen to concentrate on
the a.m. peak. If there are deficiencies in the
network, they will be demonstrated during the period
of greatest strain.

1.4 The Methodology Employed in the Transportation Model

It is beyond the scope of this work to give anything
more than a bare outline of the methodology employed
in a transportation study ₍₇₉₎.

Transportation modelling proceeds in two stages.
First, it seeks to reproduce the journeys made each
morning in a particular year (1976). To do so, it,
must identify the variables which affect present-day
travel. The basis is then laid for the second stage.
The second stage involves the forecasting of travel
in some future year (1991 was selected). By some
means or other, each of the important variables must
be projected to 1991. The result will tell you, not
only how many vehicles, how many people, and how many
journeys each day, but the origins and the destinations
of those journeys. By comparing the present day capacity
of the network, deficiencies can be identified where
demand (the number of vehicles travelling between
point A and point B) is likely to exceed supply (the
road capacity between point A and point B).

78. The matter was explored in some depth during the
course of Public Hearings. Those interested should
refer to the Public Hearings involving the Urban
Transport Study Group on the 20.11.79, 23.11.79,
4.12.79, 12.12.99, 22.1.80 and 31.1.80 (and to the
exhibits referred to in the course of those transcripts).

-128-

The 'four-step' model, as the name suggests, proceeds
in four stages. It asks, and answers, the following
questions in sequence:-

How many trips?

Where to?

By what means of travel?

By what route?

These four questions correspond with the four
of the model:-

Trip production/attraction

Trip distribution

Mode-split

Trip assignment

The road network is simplified for the purposes
the study. Local roads are usually omitted. There
are exceptions where a local road performs an import-
ant arterial function or carries a sizeable volume of
traffic. They are then included. The network
comprises some 11,000 links ₍₇₉₎. Each link represents
a particular road or portion of that road.

Having analysed what it is that generates traffic
(by means of a survey), and having proceeded through
the four-stages of the model, the simulation can be
checked against actual traffic counts.

The presence of the Harbour and the Georces River
makes the job of checking somewhat easier. If the
survey is accurate, and the procedure has been faith-
fully followed, the number of vehicles crossing the
Harbour on various bridges should more or less
approximate the number revealed by traffic counts to
have used these facilities in the year being reproduced.

79. Transcript UTSG (now STSG), 20/11/79, page 67.

-129-

The same procedure is employed elsewhere in the
Metropolitan Area. Screen-tines are drawn transect-
ing various links. The number of vehicles crossing
the screen-line is then tallied and compared to the
number revealed by traffic counts. The aim is to
reproduce the flow within plus or minus 5% ₍₈₀₎.
The experience of the State Transport Study Group
in respect of the 1976 model is described in the
following passage ₍₈₀₎:

"COMMISSIONER: We will just take the
1976 Trip Table. What happened there?
Did you in fact get that the first time?

FIELD: We got very close...the first
time.

FORD: Yes, in the first pass we looked
at two screen-lines. We looked at the
east/west screen-line, which essentially
follows the course of the Parramatta
River, the natural watercourse; it goes
below Parramatta...down to Duck Creek.
And we looked at the north/south screen-
line which is the northern railway from
Hornsby down through Thornleigh, West
Ryde, Deniston and it proceeds south
generally east of Rookwood Cemetery and
it connects to the Georges River in the
south. That effectively divides the
entire area up into four quadrants and
we looked at the crossings between each
of the quadrants. In fact, the correla-
tion was quite good."

Having successfully reproduced the 1976 traffic flow,
the Group then forecasts the growth in each one of the
variables used in that simulation. A trip-table for
1991 is the result. Unlike the 1976 table, there cannot,
in the nature of things, be a check. The modellers
must use their professional judgement. It must 'seem'
right.

We will briefly examine each of the four steps in the
model. We will identify the elements which must be
measured and fed into the computer in order to simulate
present traffic flows. In this way one can appreciate

80. Transcript UTSG (now STSG ) 20/11/79, page 74.

-130-

the enormity of the forecasting task. Every one of
the elements measured in the simulation of present
day traffic levels must be forecast fifteen years
ahead.

2. THE FOUR-STEP MODEL

2.1 Step 1: Trip Production/Attraction

2.1.1 Data Collection and Survey

In 197l the Sydney Area Transportation Study carried
out a survey. It divided the metropolitan area into
592 zones (which have since been expanded to 658 zones).
The zones formed a grid pattern covering the entire
metropolitan area. They were neither identical in
shape nor in size. Various criteria were adopted in
identifying zone boundaries. Major roads, geographical
features, or railway lines were often selected. The
selection aimed at homogeneity within each zone (whether residential, shopping, recreational or industrial).

The survey was then undertaken with a view to identify-
ing the travel characteristics of each zone. A sample
of the zone population(perhaps 2½% or even 4%) was
selected and interviewed. The aim is described as
follows ₍₈₁₎:

"The objectives of the survey and analysis
stage of the transportation planning process
are to: -

determine where journeys begin and end

determine the factors which influence
trip generation

establish the main corridors of movement.”

The sort of material covered by the survey is described
as ₍₈₂₎:

81. Introduction to Transport Planning, M.J. Brunton,
page 38.
82. Brunton ibid., page 50.

-131-

"..The household information gathered
includes the address of the dwelling
unit, the size of the household, and
the age and sex structure of the
occupants (sic); the numbers economi-
cally active, their job, and place of
work; the numbers of motor vehicles
owned; the household income; the place
of school/further education for those
occupants of school-age; and the day
and date of the journeys to be reported
i.e., normally the previous day and date."

A typical day is selected avoiding public holidays or
strikes.

2.1.2 The Trip Generation and Trip Attraction
Equations Derived by SATS

Mathematical formulae (known as models) were developed
from the survey material to disclose:

How many trips were produced by each
zone over a 24 hour period
How many trips were attracted to each
zone in the same period

The statistical technique known as Linear Regression
Analysis was used. The trips were divided into the
following categories:

Work journeys from home to an establishment
engaged in manufacturing
Work journeys from home to other forms of
work, and tertiary education
School journeys from home to various
educational institutions

The trip production equations were:

Number of manufacturing work trips from
home = 1.715 x number of manufacturing
employed residents

-132-

Number of non-manufacturing and tertiary
education trips from home = 1.715 x other
(non-manufacturing) employed residents

Trips from home to school = 1.561 x
persons aged 5 to 18 years

These equations simply tell you how many people set-off
for work from each zone each day. They do not tell
you where the jobs are. They do not tell you which of
the possible jobs each individual may take. To know
where the jobs are, it is necessary to identify the
attraction offered by each zone to the three categories
of journeys. The next step in the model (trip distribu-
tion) actually distributes the trips amongst the
attractions.

Using the same statistical technique (regression
analysis) the following attraction equations were
derived by SATS ₍₈₃₎:

Attraction of manufacturing work
from home = 1.718 x manufacturing

Attraction of other work trips (including
tertiary education trips) from home =
1.690 x other jobs + 1.314 x tertiary
enrolments

The attraction of school trips from home =
1.807 x school enrolments.

Other equations were derived for shopping trips and
social and recreational trips. They were not used.
Rather it was assumed. (again based on the survey) that
the travel covered by these equations (and by the
commercial vehicle equations to which we will refer)
represented approximately 90% of the peak period travel.
Very little shopping is done in the peak because shops
are usually not open until after 9 a.m. The procedure

83. SATS Volume 2, Chapter II, page 9, Table 2.1

-133-

adopted is explained in the following passage ₍₈₄₎:

"FIELD: Something like 80% or 90%
of all travel in the peak period is
explained by those three purposes
(i.e., manufacturing trips, other
trips and school trips) and when
eventually we talk about the mode
split process, that 90% of travel..
is expanded to represent 100%."

Employing the same technique, the following commercial
vehicle trip generation equations were derived ₍₈₅₎:

Commercial vehicle (light vehicle trips) =
55.9 x retail acres + .762 x manufacturing
jobs + .395 x other jobs + 711 x families

Commercial vehicle (heavy vehicle trips) =
2.1 x industrial acres + .375 x manufacturing
jobs + .165 x family + 22.1 x (retail acres
+ office/commercial areas)

There is a further equation for taxis. We gather it was
not used by the Study Group.

2.1.3 The Information Required to Calculate the Number
of Trips

It will be appreciated from these equations that an
enormous body of data is required in order to predict
the traffic generated throughout Sydney. Given the
breadth and complexity of Sydney (or any city) it would
be surprising if it were otherwise.

A convenient summary of the categories of information
appears in a document provided to the Inquiry by the
State Transport Study Group. The document ₍₈₆₎ is
apparently intended to provide newcomers to the Group
with some sort of introductory statement on the
workings of the model. It reads in part ₍₈₇₎:

84. UTSG (now STSG) Transcript 20/11/79, page 72.
85. SATS Volume 2, Chapter II, page 9, Table 2.2.
86. Exhibit 34, dated May, 1975.
87. Exhibit 34, page 3.

-134-

“If using the SATS equations the following
zone or data is required:

ZONE DATA: for (home to) work and
(home to) school trips

manufacturing employed residents
other employed residents

persons aged 5 to 18

manufacturing jobs

other jobs

tertiary enrolments , full-time
by age

school enrolments by age

ZONE DATA: for commercial and external
cordon vehicle trips

retail area - site area not
floor space

office commercial area - site
area not floor space

industrial area

manufacturing jobs

other jobs

families."

This catalogue of variables influencing travel
ultimately must be predicted for the year 1991. The
task is enormous, and it grows, as a progression
through the four-stages of the model demonstrates.

2.1.4 The Compilation of Data by Zone

The zones are unique to the Study Group. They have
been devised for the purposes of the traffic model.
In the nature of things the data received by the
Study Group is expressed in terms of local government
areas or some other administrative sub_division. It
has to be broken down into zones.

The difficulty can be explained by an example.
Estimates of population are provided from time to
time by the planning and Environment Commission.
These estimates are given for local government areas.
A further sub-model has been devised known as the

-135-

Planning Parameter Allocation Model (PAM) for this
purpose. The aim of the sub-model is described in
the following paragraph ₍₈₈₎:

"The Planning Parameter Allocation Model
attempts to allocate, on a rational basis,
a given planning strategy from the strategic
level to a zonal level...the model can
allocate strategic level forecasts of
population and employment through traffic
zones on the basis of each zone’s charac-
teristics and its capacity to absorb change."

It will be appreciated from this description that it
is necessary to make further assumptions, and intro-
duce further simplifications, even when the data has
been amassed.

2.2 Step 2: Trip Distribution

It will be remembered that the sequence of questions
is:

How many trips?

Where to?

By what mode?

By what route?

Trip distribution is the second phase. It seeks to
answer the question: where to? Having determined
the number of trips produced (or attracted) by the
various zones, they must then be distributed amongst
those zones.

There are several possibilities. The one most
favoured (and the one used in Sydney) is known as
the Gravity Model. Its formulation resembles
Newton’s formula for gravity. It rests upon a
concept which is familiar to all of us. The
probability of a person working in a given place of
employment declines the further he or she lives from
that place.

88. Part of Exhibit 43, Article "Planning Parameter
Allocation Model", in the Greater London Intelli-
gence Quarterly, No. 29, December, 1974, page 56.

-136-

Mathematically the formula is expressed in the
following way ₍₈₉₎:

The same formula can be represented graphically in
the following way (to take the example of work trips
from home to manufacturing employment) ₍₉₀₎:

FIGURE 7.

WORK JOURNEYS TO
MANUFACTURING JOBS FROM HOME

89. SATS Volume 2, Chapter II, page 17.
90. SATS Volume 2, Chapter II, page 20.

-137-

Members of the public need not concern themselves
with the mathematics. It may or may not be
comprehensible to them. What is comprehensible,
and what is of concern, is the nature of the data
which must be assembled to use the formula. It is
this data, as we have said before, which must then
be predicted fifteen years on to 1991. Members of
the public can well appreciate the difficulty of
that task.

It will be seen from the graph that the gravity
model requires the calculation of 'composite cost
units'. The formula appears in Exhibit 35. It is
not reproduced here. It is sufficient if we simply
list some of the ingredients which must be separately
calculated:

It requires details of car ownership
and the percentage of households with
one or more cars available
It requires a calculation of highway
costs which, in turn, are derived by
estimating the time it takes to get
from one point to another and the cost
of travel. The cost of travel
incorporates petrol prices etc.
It requires an estimate of the public
transport cost which, in turn, involves
an estimate of the time it takes to get
from one place to another, and the cost
in terms of fares

It will be appreciated that the difficulties in predict-
ting these elements fifteen years ahead are formidable
indeed.

2.3 Step 3: Modal Split

2.3.1 Introduction

At the end of the second stage a number of trip
tables are produced. A trip table is no more than
a matrix of trips between zones. You now know how

-138-

many people (or commercial vehicles) are journeying
from one zone to another. The number of vehicles
(as opposed to the number of people) making that
journey will depend upon:

How many use public transport

How many occupants in each car

For all but commercial vehicles, therefore, it is
necessary to go through a further phase. The number
of public transport users is calculated by a
separate procedure. For those using a
a separate model (i.e., a mathematical formula)
disclosing the likely vehicle occupancy between one
zone and another.

2.3.2 The Methodology Employed in Determining
Public Transport Users

The methodology employed by the State Transport
Study Group was developed by SATS. It is based
therefore, upon the survey made in 1971. That
survey endeavoured to isolate those matters which
appeared to influence the choice of mode (whether
public transport or car). In broad terms the
factors were:

1.      Whether a car was available or more
   than one car

2.      The time and costs of travel by public
   transport and by car

3.      Employment density (the number of
   employees per acre) ₍₉₁₎

The time and cost of travel is itself complex. It
requires a mass of data. In calculating the cost of
travel by public transport, the total door-to-door
travel time must be obtained by assessing each
component of the journey, namely ₍₉₂₎:

91. See generally SATS Volume 2, Chapter II, pages 27-36.
92. SATS ibid., page 29.

-139-

Access and egress time, which is the
time taken to walk, drive or to be
driven to and from the public transport
station or bus stop
Waiting time; i.e., the time spent waiting
for public transport or in transferring
from one vehicle to another. Bus and
railway time tables are used and the
waiting time is estimated to be half the
head-way. The head-way is the interval
between services on that line or through
that area
The in-vehicle time
The public transport journey fares. The
prediction of this aspect is made rather
more easy by the recent announcement of
the State Government that fares will be
indexed to the inflation rate

By means of a mathematical formula a calculation can
be made of what is termed the 'public transport
disutility'.

The formula for this calculation, and its companion,
highway disutility, is reproduced below from the SATS
Report ₍₉₃₎:

The State Transport Study Group provided the Inquiry
with an outline of the steps taken in applying the
mode split model. There are eight steps ₍₉₄₎. We
will not repeat that information in this Report. It
is available for those who have an interest. Our
purpose, as we have said, is to identify those
matters which must be predicted.

93. SATS Volume 2, Chapter II, page 31. See next page 140.
94. Transcript UTSG (now STSG) 20/11/79, pages 51-66.

-140-

2.4 Step 4: Trip Assignment

Three questions have now been answered:

How many trips?

Where to?

By what mode?

We must now address the fourth question:

What route?

We know from the second stage (the trip distribution
stage) how many commercial vehicles. We know at the
end of the mode split stage how many ordinary
vehicles. These vehicles must now be 'loaded’ onto
the network.

-141-

There are several techniques. We will begin by
describing the most simple. We will then refer to
the transcript for a description of the technique
employed by the State Transport Study Group in
developing the model used in the present Inquiry.

The most simple technique is called the 'all-or-
nothing' assignment. At the gravity model stage
(the second stage) it is necessary to work out
minimum time and distance paths between zones so
that trips can be distributed. The same informa-
tion is used in the assignment process. The
procedure involves the following steps:

1. Defining the highway network

2. The determination of the minimum
‘cost' paths between zones. ₍₉₅₎
The minimum cost is based upon:

the minimum distance paths
between zones

or the minimum time paths between
zones

or some combination (with an
appropriate weighting in favour
of time or distance)

This stage is often referred to as
'building trees’.

3. The assignment of all vehicles from
   each zone to every other zone by the
   appropriate minimum cost path.

4. The aggregation of total flows on
   each link in the defined network.

The technique has certain disadvantages. They are
described in the following passage ₍₉₆₎:

95. See UTSG (now STSG) transcript 20/11/79, pages
67-68 .

96. Bruton op.cit page 144.

-142-

"One major drawback is that the
technique takes no account of
increasing congestion associated
with increased volumes, and assigns
too many vehicles to the better
routes as travel times on these
routes will be better than on the
multi-purpose streets. . . .Small
differences in journey times by
different routes between the same
origin and destination can bring
about unrealistic journey paths
when the all-or-nothing assignment
is used."

The alternative method of assignment endeavours to
meet these objections. It takes account of the
congestion which is likely to be encountered upon
the more popular routes. That congestion will
operate to lower journey times. The computer is
able to adjust the speed of vehicles according to
the loading of the link. The procedure is described
by Bruton as follows ₍₉₇₎:

"The first stage of the capacity
restraint assignment involves the
building of minimum path trees in
exactly the same way they are built
for the all-or-nothing technique.
Traffic is then assigned to these
minimum paths, but as the assigned
volume on each link approaches the
practical capacity, the computer,
through an iterative procedure in
which loaded link information is used
as the feed-back to the tree-building
process, automatically lowers the
assumed speeds on the effected links,
thereby making these links less
attractive to traffic."

The particular technique used by the Study Group
is described in the following evidence (98):

"FIELD: We use a technique called a
stochastic path building technique,
which says that 90% of the people
might take the shortest path but 10%
might take paths which are a certain
amount outside that shortest path..
People don't always perceive (costs)
in the same terms.

97. Bruton ibid., page 145.
98. Transcript UTSG (now STSG) 20/11/79, page 68.

-143-

COMMISSIONER: This is the one where
you do it, 10% at a time?

ANSWER: That’s right, in fact we use
20 increments. What it means is that
you have 658 zones. The programme
starts at that origin. (It) builds a
set of paths to all 658 (zones). They
represent not just one path between
each origin and destination but maybe
three.”

The traffic between one zone and another is, by this
means, loaded onto the network. Mr. Field explains ₍₉₉₎:

"FIELD: So the speed perceived by the
first person is not the same, if you
like, as perceived by the second
increment. Curves are again used to
relate the number of vehicles on a
road with the capacity and the type
of road...Every increment perceives
the congestion caused (by previous
increments)."

The view was then expressed that the assignment technique
was highly reliable ₍₁₀₀₎.

2.5 Forecasting with the Four-Step Model

One can only gasp in amazement at the miraculous way
in which this elaborate procedure manages to reproduce
the traffic using the Sydney Metropolitan Network
within plus or minus 5% of the actual traffic flows.
The technology is truly amazing.

But we must not allow our admiration for the extra-
ordinary expertise and technical competence displayed
in such an exercise to blind us to the enormity of
the quite separate task of then predicting 15 years
ahead each one of the ingredients important to the mix.

A useful summary of the most important factors which
must be predicted is furnished by the State Transport
Study Group itself. It said ₍₁₎:

99. Transcript UTSG (now STSG) 20/11/79, page 68.
100. Transcript UTSG 20/11/79, pages 78 & 80. See,
     however, page [sic]
1.   UTSG (now STSG) A revision of the SATS year 2000
     Travel Forecast, July 1975. Page 7.

-144-

"..There are difficulties and problems
in the prediction of 16 factors (used
in the trip generation equations) for
each of about 600 zones, the main ones
being:

a) the overall study area population..

b) distribution of population, which
is related to densities, lifestyles,
market supply and demand etc.

c) the distribution of employment,
especially that commercial employ-
ment, whose location is (in theory)
subject to planning policy control."

d) the distribution of other activities
(schools, recreation, shopping)
which cause trips to be made in
the a.m. peak period; and

e) car ownership levels (related to the
desirability of private cars), car
running costs (especially fuel) and
hence car usage. "

We will briefly examine some of the evidence relating
to some of these issues. In many cases we are not in
a position to say that the assumptions made are wrong.
In every case it is clear that the Study Group has
gone to great lengths to discover the most recent and
reliable information. Even if the information fed
into the prediction was absolutely accurate in every
detail, the problems of forecasting brought about by
sudden changes in trends (including those induced by
policy initiatives) must make one approach any
prediction, 15 years ahead, with some circumspection.

3. LIMITATIONS OF THE FOUR-STEP MODEL

3.1 Introduction

The distinction between the use of the model as a
predictive device, and its use as a mean of fashioning
solutions to present day problems, must be constantly
borne in mind. It is used by the State Transport Study
Group in both roles. The following was said during the
course of public hearings ₍₂₎:

2. Transcript UTSG (now STSG) 20/11/79, page 8.

-145-

“COMMISSIONER: I am still having some
trouble in discovering why it is that
you go to the enormous trouble that
you do to construct the 1976 base year.
I can understand in the case of 1971.

FIELD: In practice, the largest thrust
of URTAC on policy is in fact to short-
term; today’s problems. One of the things
they asked us to analyse are the problems
of today. To do that, one tool we use is,
in fact, the trip table for 1976 which
provides information that is not available
from any other source. It provides a
consistent basis, Sydney-wide, of travel
patterns.."

Many of the limitations arise when the model is
employed as a predictive tool. There are those (and
Professor W.R. Blunden is amongst them) who maintain
that the primary function of the model is to assist
analysis and synthesis not prediction. What would be
the effect of providing a link in a particular place?
Would a policy dedicated to a particular land use
pattern minimise travel? In a paper which is annexed
to this chapter professor Blunden says ₍₃₎:

"The major misconception on the application
of transport or land/use transport models
is, however, the reliance placed on them
as predictors of the 20 year future. To
be useful as a rational analytical tool
their role must be limited to one of
analysis and synthesis.”

Later he says ₍₄₎:

“I would repeat here that the role of the
model is for analyses and syntheses and
not to provide an elaborate surrogate
for the crystal ball."

The State Transport Study Group is more than aware of
the model's limitations, especially in the realm of
prediction. Its view, indeed, is not greatly different
from that of Professor Blunden. The following words

3. Annexure 1: Transportation Modellinq – An
Esoteric Exercise, W.R. Blunden, page 337 of this
Report.

4. ibid., 337.

-146-

were written with the Sydney Area Transportation Study
in mind (5):

"Most of the above problems (which were
problems of prediction) arise from the
adoption of the long range planning
horizon - anything can happen in 30
years and this means that forecasts
should not be used to develop a fixed,
definitive plan. Rather, the forecast
should be treated from a strategic view-
point, giving a broad indication of peak
period travel volumes for a given set of
assumptions, and, in this case, seeing what
general effect changes in one of the
assumptions (the population level and
distribution) might have on the forecast..
The primary use of this report’s forecast
is to indicate in broad terms how a
particular change in Sydney's development
pattern could alleviate or worsen its
transport problems..."

3.2 The Model Assumes Tomorrow Will Be a Bigger Version of
Today

It will be remembered that a survey was carried out
by SATS in 1971. The survey aimed, first, at identifying
the important factors influencing travel habits. More
than this it aimed, secondly, at constructing equations
(by means of regression analysis) which could then be
used as a means of predicting the capacity of each
zone to generate (or attract) travel in the future.

Regression analysis does not uncover the causes of
travel. It simply identifies matters which are
statistically significant at one point in time. Without
an appreciation of the causal relationships involved it
is difficult to say with confidence that precisely the
same factors will operate in 20 years’ time. One
commentator, Blackshaw, says this ₍₆₎:

"..The procedures are mechanistic, being
solely concerned with reproducing exist-
ing travel patterns. Because the models
concentrate on variables which give a
good fit, rather than those which present
causal relationships of behavioural or
policy significance, they are unlikely to

5. UTSG (now STSG) A Revision of the SATS year 2000
Travel Forecasts, July, 1975, page 8.

6. The SATS Study - An Economic Review, P.W. Blackshaw,
page 59.

-147 -

yield even tolerably accurate forecasts
of future travel as the city grows, as
relative locations of residences and
work places change, and as perhaps the
relative attractiveness of competing
modes changes."

One rather gathers the State Transport Study Group
would not disagree. Commenting upon another model
which employs the same technique, regression analysis,
it says ₍₇₎:

"..The stability of the regression
coefficients over time and between study
regions is tenuous. There is an implicit
onus on the researcher to incorporate
meaningful relationships and not simply
those selected on the basis high correla-
tion.”

And again ₍₈₎:

"As the model was calibrated at one point
in time there is no built-in mechanism
to analyse the way components may change
over time. In the absence of knowledge
regarding the true causative factors
determining density it is tenuous indeed
to assume a constant quasi-causal rela-
tionship existing over time, even if high
correlations had been found in the base
year."

The time horizon chosen for the present Inquiry was
15 years (1976-1991). However, the same SATS
equations, derived from the 1971 survey, were used.
Essentially, therefore, the constancy in time
postulated is between 1971 and 1991: a period of
20 years.

3.3 Performance of Other Similar Traffic Models

What has been the track record for models of this
sort? The evidence is mixed.

7. UTSG (now STSG) Land Use/Transport Interaction:
A Working Paper, page 59.

8. UTSG ibid., page 86.

-l48-

Blackshaw refers to one study nine years after
certain predictions had been made. He says (9):

"Ashford and Holloway (November 1972)
applied a trip production model estimated
from data from the 1958 Pittsburg Area
Transportation Study to 1967 socio-
economic data and compared the results
with actual 1967 trips" They found
'wide divergences' of up to 30%, with
the important 'home based work trip'
category over estimated by 15%.”

The Leitch Committee in the United Kingdom (l0) was
concerned with the accuracy of traffic forecasts. As
part of its investigation it arranged for comparisons
to be made between forecast traffic flows and the
traffic actually using the facility once it was built.
It concluded ₍₁₁₎:

"..These comparative figures seem to
support the general conclusion that
there has been a tendency in the past
to over-predict traffic, in certain
cases significantly."

The Study Group furnished the Inquiry with certain
American Literature reporting on comparisons between
transportation forecasts (made by employing a model
similar to that used in Sydney) with observed traffic
flows. In respect of one (for Indianapolis) the
following emerged ₍₁₂₎:

"The result indicated that the household
models based on the 1964 data could
successfully predict household travel
reported by the same households in 1971.” ₍₁₃₎

9. Blackshaw ibid., page 60.

10. Report of the Advisory Committee of Trunk Road
Assessment, October 1977.

11. The Leitch Report ibid., page 76.

12. Temporal Stability of Trip Generation Relations,
Kannal and Heatherington in Highway Research
Record No. 472 at page 17.

13. See also ‘Time Stability Analysis of Trip Generation
    and Predistribution Modal-Choice Models’ Smith and
    Cleveland in Transportation Research Record No. 569
    at page 76.

-149-

The most recent analysis was conducted by the
Institute of Transportation Engineers in the United
States. Its findings were reported in its Journal
of February 1980 ₍₁₄₎. It examined transportation
studies from 15 cities. It compared the forecasts
made by the model with the traffic observed in the
forecast year. Its conclusion is calculated to
induce elation and despair simultaneously. It says ₍₁₅₎:

"Often the conclusions drawn from one
forecast contradict those based on
another forecast. For example, there
is evidence that the longer the time
period included in the forecast, the
more the forecast will deviate from
observed values. However, the relatively
short-term (6 years) population and
employment forecasts made in the Puget
Sound area had larger errors than longer
11 year forecasts of same.”

And further ₍₁₅₎:

"The more items that are forecast increase
the vulnerability of those responsible for
making them because there are more chances
for the forecasts to 'miss'. However, if
more items are forecast and the inter-
relationships among them are kept in proper
perspective, the possibility of compensatory
errors may be increased when travel demands
are estimated."

The article concludes by acknowledging that there have
seen 'significant deviations in past forecasting urban
area activity', but goes on ₍₁₅₎:

"The ultimate design of the facility is
usually based upon a saturation or mature
situation. As such, the facility should
be designed to accommodate this demand when
ever it occurs, be it five years early or
five years later than originally forecast.
There may be some short-term misallocation
of resources for forecasts which are not
realised in a timely manner..."

14. Evaluation of the Accuracy of Past Urban Transport-
ation Forecasts I.T.E. Journal February 1980, page 24.
15. ibid., page 34.

-150-

Closer to home, it is apparent that the Sydney Area
Transportation Study over-estimated traffic demand.
Within one year of completion it was revised by the
Study Group ₍₁₅₎. That revision will require even
further adjustment to take account of more recent
population figures which suggest an even slower rate
of growth ₍₁₇₎.

3.4 The Model Assumes a Fixed Trip Matrix

3.4.1 Trip Suppression Through Congestion

The following assumption is made (18):

"Congestion on the road system, region-
wide, (will) not (change) substantially.”

The traffic models were devised at a time when neither
money nor philosophy were serious impediments to road
building. It was assumed as a matter of philosophy
that free flowing conditions were appropriate though-
out an urban area. It was also assumed (and with some
justification) that there would be sufficient money
available to achieve that end. There have been changes.
We have already discussed congestion. We have suggested
that in certain situations (especially near the Central
Business District) congestion may be inevitable. It
ought not be corrected lest traffic be encouraged in a
direction in which it should be discouraged.

It, is inevitable, therefore, that a policy dedicated
to the selective elimination of congestion must
suppress demand in those areas where the planners
choose to allow congestion to remain. The tightly
controlled road budget, moreover, must ensure tha
there are worthy projects which must be put to one
side until funds become available, The presence of
congestion, and the suppression of demand, are inconsis-
tent with the assumptions made by the model.

16. A Revision of the SATS year 2000 Travel Forecast,
July, 1975.

17. Technical Bulletin No. 8 (August 1976) and more
recently the PEC publication 'Population Projections
for New South Wales 1976-2001' February, 1976.

18. D.M.R. Submission S.K/C 340 Transport and Economic
Analysis September 1979, page 3.

-151-

The State Transport Study Group put the matter in
this way ₍₁₉₎:

"FIELD: One would like to see trip production
and attraction related to system accessibility.

COMMISSIONER: Sorry, what does that mean?

FIELD: That is a highly congested system is
not accessible and hence would be likely to
produce less trips. A freely flowing system
is very accessible and may produce more trips.

COMMISSIONER: So the extent to which demand
is being suppressed by congestion is not
taken into account?

ANSWER: No.."

This limitation was acknowledged by the Sydney Area
Transportation Study ₍₂₀₎.

What are the consequences? It is suggested by some
that the use of a 'four step' model on a congested
road network is simply inappropriate ₍₂₁₎. Others
suggest the model may significantly over-estimate
demand ₍₂₂₎.

3.4.2 Traffic Generated by the Facility

We have referred already to the tendency of a new
facility to generate traffic ₍₂₃₎. The model takes
no account or this tendency. SATS puts it in this
way₍₂₄₎:

"The model does not make any allowance
for the additional trips which might
be made in a superior quality of travel."

The Miami example is a good illustration ₍₂₅₎. The
model carefully worked out the potential demand for

19. Transcript UTSG (now STSG) 20/11/79, pages 14-15.

20. SATS Volume 2, Chapter I, page 1.

21. New Technique for Transport Systems Analysis
R.W. Morris, page 9.

22. The Urban Transportation Planning Process O.E.C.D.,
page 30.

23. See page 55.

24. SATS Volume 2, Chapter I, page l.

25. See page 57.

-152-

the year 2000. The facilities were built in the
space of 10 years. Their very presence generated
traffic. Suddenly people thought ‘how nice it would
be to live in a better part of Florida’. People
sought to take advantage of the opportunities created
by the facilities.

3.5 The Model Assumes a Fixed Land Use Plan

We have stressed already the interactive nature of
land use and transportation planning. The land use
pattern will be different if facilities are built
than if they are not built. It will be different
if they are built here rather than there.

The modelling process is generally not used as an
analytical tool to assess the desirability of land
use and transportation combinations, to bring them
into a more favourable equilibrium state.

Professor W.R. Blunden puts the matter in this was (26):

"..The standard urban transport planning
packages currently in use do not take
account of the land use factors except
as exogenous input variables. In major
studies so much time and effort goes
into taking out one transport solution
based on given land use inputs that a
re-run with a new land use input is
generally disregarded until a new
"transportation study" is commissioned.”

The Study Group is certainly not oblivious to the
problem. They have said ₍₂₇₎:

"It would thus appear more rewarding to
change our methodology and promote a
land use/transportation interactable
approach at the grass-roots i.e., formu-
lation of ‘land use/transport system'
alternatives, rather than transport
systems to 'fit' a predicted, inflexible
land use pattern."

26. Annexure 1: W.F Blunden ibid., page 337 of this
Report.

27. UTSG (now STSG) Land Use/transport Interaction:
A Working Paper, page 87.

-153-

The transportation models were devised in the United
States of America. They have been used in a great
many transportation studies in that country. The
accumulated wisdom of that experience is worthy of
especial attention. The current practice in the
United States is described by K.W. Dobinson in these
words ₍₂₈₎:

"The U.S. Department of Transportation
is conscious of the trends outlines
(i.e. the change in lifestyle and sprawl
brought about by improved transportation
facilities, whether road or rail). It is
also aware of the deficiencies in the
conventional transportation model to
determine the city’s transport needs,
which model is still used for this
purpose in Sydney. This is because the
conventional transportation model assumes
a static relationship between the spatial
location population, job location and the
transport system. Urban structure is
much more complex and any significant
change in structure, as represented by a
major transport facility, will have a
variety of effects - geographic, social,
economic - to mention a few. This was
the case in San Francisco, Toronto and
Miami where the transport system designed
to meet demand did not do so but simply
caused a spread in the spatial location
of the population.

(emphasis added)

The dilemma can be illustrated in this way. If the
objective is to minimise travel and to promote regional
self-containment (as we have suggested) it is desirable that industry should move to those areas where jobs are
in short supply. Yet assume (as an illustration) that
the trend was otherwise. The trend may be for jobs to
remain concentrated in the central Industrial Area.
Diligently reading those signs, the transport modeller
may predict a continuation of that trend. That
prediction will flow through the traffic model and
produce certain traffic loadings. Those traffic flows
may suggest the road network in the vicinity of the
Central Industrial Area is severely overloaded. That

28. Car Cult Country K.W. Dobinson, page 12.

-152-

overloading may be used to promote road works in
the Central Industrial Area. Yet those road works,
and the eradication of congestion, may be the very
thing that keeps industry in the Central Industrial
Area and prevents it moving out. In short, the
model may suggest precisely the wrong solution.

The following was said by a member of the Study
Group ₍₂₉₎:

"FIELD: ..But as congestion occurs people
will change, or tend to change their travel
patterns or their location decisions, people
and the economy.

COMMISSIONER: In some cases that is desirable.

FIELD: Oh yes, absolutely. It is suggested
it’s happening all the time in Sydney. That’s
because Sydney isn't building heavily in a
road system. It means a system just doesn’t
grind to a halt. Things must change. So one
can say that the development of the West and
the outer fringe areas is a direct response
to the congestion of the inner area. And
certainly, possibly the job shift out of the
(Central Industrial Area) could be possibly
due to congestion in the area. It’s one of
many effects."

3.6 The Model Ignores the Capacity of Policy to Alter
Trends

The model is concerned to discern trends. In the
nature of things it ignores the capacity of planners
to alter trends. In a Review of Transportation
Planninq in Australia, NAASRA put the point in this
way ₍₃₀₎:

"While trend extrapolations may be reliable
in the short term, they may imply that
existing trends are either desirable or
immutable, and may ignore the possibility
of altering trends by various discretion-
ary policy measures."

The model, for instance, simply assumes car ownership
will increase as it has in the past. The Leitch

29. Transcript UTSG (now STSG) 4/12/79, page 14.

30. Review of Transportation Planning in Australia,
National Association of Australian State Road
Authority 1977, page 8.

-155-

Committee said this ₍₃₁₎:

"A further criticism that has been put
to use is that. The forecasts take a
passive attitude towards future policy
changes. That is to say, on the basis
of past growth the Department has
extrapolated car ownership and use
into the future and does not consider
whether the levels of car ownership
and use implied by this forecast are
desirable.. Commentators believe variously
that the Government should decide whether
the levels of car ownership implied by
the forecast for the future are desirable
and, if not, should take steps to reduce
them; that a policy of encouragement of
other modes should be pursued to slow
down the rate of increase of car owner-
ship implied by the forecast; that a
policy of traffic management in the
wider sense, should be followed."

(emphasis added)

There was a useful discussion of this aspect in the
public hearings. On the question of policy decisions
affecting land use, the following was said ₍₃₂₎:

"COMMISSIONER: So that in terms of land
use, of necessity, you must project the
situation in which there will be no
policy changes between the present time,
the date of your calibration, and the
forecast year. Is that right?

LEAVENS: I would say in broad terms, yes.

COMMISSIONER: Do you take into account
some general trends that you feel are in
the offing?

LEAVENS: Again we are guided by the
Planning and Environment Commission's
assessment as to how population will
wish to distribute to the major sectors
of the metropolitan area, but one thing
we are fairly confident of is that the
release areas identified in the Sydney
Region Outline plan will remain roughly
the same."

The matter must be kept in perspective. The Study
Group said this ₍₃₃₎:

31. Report of the Advisory Committee on Trunk Road
Assessment (Leitch Committee), page 5.
32. Transcript UTSG (now STSG) 23/11/79, page 29.
33. Transcript UTSG (now STSG) 23/11/79, pages 37-38.

-l56-

"LEAVENS: I would have to agree that
policy can make dramatic changes in
direction..One thing we are always
dealing with is a lot of sunk
investment in the urban fabric in
the jargon of town planning. If
you drive through the streets you
see a lot of houses that were built
in 1900 and you see a lot of houses
that were built in 1940. They have
not been torn down. The actual
capacity of residential tenements in
most of the urban area will not change
in 15 years... "

It could not be suggested, in other words, that
policy will dramatically effect the entire metropol-
litan area. Rather, it may inhibit or induce changes
in the industrial areas, or it may influence the way
in which the urban fringe develops. Both those
matters are important in this Inquiry. One issue
is whether the construction of the South Western
Freeway or the Cooks River Option will make the
Central Industrial Area more attractive to industry.
Is it better to leave it as it is in the hope that
industry will decide to relocate in the West or South-
West? Another issue is the growth which can be
expected in Campbelltown (the Macarthur Growth
Area) and the Western region. Policy may influence
that growth.

3.7 Insensitivity of the Model to Fuel Shortage

The following assumptions were made in respect of
fuel ₍₃₄₎:

“Usage of vehicles in peak periods
(will) not (be) unduly constrained
by the availability or the price of
fuel.

Petrol prices being doubled 1976
(sic) levels (in real terms)."

34. DMR Submission S.K/C 340 Transport and Economic
Analysis, page 3.

It was assumed, in fact, that petrol prices would
be double the 1971 level in real terms. Between
1971 and 1976 the price of petrol (in real terms) fell
by approximately 6%. It has risen sharply since
that time.

The limitation of the model is described by the
Department of Main Roads in these terms ₍₃₅₎:

"Three factors of particular interest in
forecasting the amount of future travel
by motor vehicle at the present time are
the leve1 of service provided by the
transport system, the price of fuel and
the availability of fuel.

The trip generation phase of the ‘four-
step' model, where the total amount of
travel is forecast, is not sensitive to
these three factors. Models incorporating
one or more of these features have been
proposed. They are not yet well-tested;
nor have they been calibrated for Sydney.
Consequently, the sensitivity of trip
generation to these factors has to be
assessed on a subjective basis and from
experience.”

The price of fuel finds its way into the model in
determining the mode sp1it. It will be remembered
that the choice between public transport and car
depends, in part, upon the cost of each. The cost
of car journey, in turn, depends upon the price of
fuel.

This is a grave shortcoming of the model as the study
Group is only too willing to acknowledge ₍₃₆₎:

"COMMISSIONER: As a matter of commonsense,
would not the supply of fuel be an import-
ant factor which one would think...

FIELD: It would be, but the supply of
fuel poses an enormous modelling-problem.
It should be taken into account and
probably in our next set of models will
have to be. But it is an enormous problem
presently as you say, it is not taken into
account in these models.”

35. DMR submission S.K/C 340 Transport and Economic
Analysis, page 3.
36. Transcript UTSG (now STSG) 2O/11/79, page 41.

-158-

We will briefly examine the energy issue later in
this chapter. The Study Group took the view that
the assumption made in 1976 (that fuel prices would
double in real terms by 1991) is still realistic
in 1980, notwithstanding the events in between . It
suggested that fuel price was far less important
than fuel supply because of the inelastic nature of
the demand for travel.

The public's participation in this Inquiry has been
of immense benefit. It has identified matters of
importance to the community. It has provided an
insight into community attitudes on a number of
issues. It has given the Inquiry the benefit of
accumulated local knowledge. The Inquiry would
have been the poorer without it.

The public's confidence in an inquiry is a fragile
thing. There is the danger that the intricacy and
complexity of the model will baffle the public into
silence. Because of a tendency, bred in all of us,
to stand in awe of things we do not comprehend, the
public may be discouraged from participation if the
impression is given that its contribution can only
take the matter so far, and that the issue will be
determined by reference to something far more
'sophisticated'.

The Leitch Committee outlined certain criteria which
any method of assessment should meet. It said (37):

"In our view the assessment should seek
as far as possible to meet the following
criteria:

(a) It should be generally comprehensible
to the public and command their respect.

(b) The public should be able to identify
how different groups of individuals
would be affected by the scheme.

37. Leitch Committee op.cit page 91.

-159-

(c) It should be comprehensive in
    terms of the different kinds of
    effects of the road scheme;
(d) It should not be expensive to
    use; and
(e) It should balance costs and
    benefits (however described) in
    a rational manner. "

These are sensible criteria. Applying them to the
modelling process the Committee said ₍₃₈₎:

"The first criterion is that the assess-
ment should be generally comprehensible
to the public and should command their
respect. The evidence which has been
presented to us shows quite clearly that
this is not currently the case. Indeed,
we have seen that COBA (the acronym for
a similar transport planning model in the
United Kingdom) is regarded by some as an
attempt to 'bamboozle’ the public and as
some sort of 'Solomon Machine'."

Given the importance of public participation, it has
even been suggested that traffic modelling is inappro-
priate. Ivan Illich expressed the idea in these words ₍₃₉₎:

"Participatory democracy demands low
energy technology."

Modelling is the reverse ₍₄₀₎:

"Modelling is a high energy activity. If
the sum of energy required to produce
and utilise a model is calculated it will
be obvious that it is far in excess of
the resources of the majority of those
groups and individuals outside the planning
process who are currently demanding partici-
pation, or whose contribution to planning
has been accepted at a formal political
level as desirable."

38. Leitch Committee ibid., page 91.
39. Energy and Equity Ivan Illich 1974, page 24.
40. Transportation Studies & Normatic Social Planning
Patric Healey & John Stanley Volume 13 Planning
Institute Journal 1975, page 84.

-160-

What should be done? This Inquiry does not suggest
that modelling should be abandoned. It is a valuable
analytical tool. It does mean, however, that an attempt
should be made to demystify the procedures followed.
The modeller should state the assumptions made in the
course of constructing the model, and the basis for
those assumptions.

It should not be thought that this Inquiry is, by
these words, expressing a criticism of the Study
Group. It is not. This Inquiry is the first of its
kind. Hitherto, the Study Group has performed its
task behind closed doors. Those to whom it presented
its material were, by and large, familiar with its
method. There was neither the time nor the need to
'spell out’ exactly what it was doing.

That has changed. The community is quite capable of
examining critically the assumptions. It is capable
of making a contribution to the debate. This Inquiry
will be the richer for that contribution.

3.9 The Possibility of Significant Change

The SATS Study attempted to make predictions for the
year 2000 (a 30 year period). It provoked the comment
from the Study Group to which we have referred:
'anything can happen in 30 years' ₍₄₁₎.

The period has been halved. A 15 year time horizon
is now used. Even that may be ambitious. A lot can
happen in 15 years. There may be changes in produc-
tivity by the introduction of technology (such as
computer chips) which may have profound repercussions
on the journey to work. There was a time when the
current level of unemployment (about 6%) was considered
a temporary aberration. It now seems probable that
unemployment at this level (or even greater) will
persist for some time to come. One does not have to

41. UTSG (now STSG) A Revision of the SATS Year 2000
Travel Forecast, page 8.

-161-

resort to science fiction to read accounts which
foretell the possibility of society sharing limited
jobs between a greater number of individuals, to
avoid massive unemployment. We do not suggest
these things will come to pass. We do not know.
They may. We simply repeat what is obvious: that
the future defies the prescience of ordinary mortals.

3.10 The Move Towards Simpler Models

Blackshaw offered the following comment upon the
methodology of the Sydney Area Transportation Study
(which employed the 'four step' model after an
elaborate survey) ₍₄₂₎:

"Because of their complexity the travel
model..are very expensive to run, thereby
drastically restricting the number of
alternatives that can be compared. The
client would be better served if a survey
(and subsequent demand modelling) was
coarser...”

NAASRA in 1977 made the following recommendation ₍₄₃₎:

"The transportation process should: ..

place less emphasis on developing
complex models and modelling techniques
(which, in the past, have taken much
time and effort), and in view of the
inherent inaccuracies of input data, and
of the many factors that make future
estimates of land use and of travel
so approximate, broader overall models
should be developed and be used as far
as possible - particularly for rapid
assessment at broad levels."

Professor W.R. Blunden in the paper annexed says
much the same thing ₍₄₄₎.

K.W. Dobinson describes the American practice ₍₄₅₎:

42. Blackshaw ibid., page 295.
43. Review of Transportation Planning in Australia,
NAASRA 1977, page 3.
44. Annexure 1: W.R. Blunden ibid., page 340 of this Report.
45. Car Cult Country, K.W. Dobinson, page 12.

-162-

"The experience of cumbersome planning
processes that have not provided an
answer to city planning or to the
transport needs of the city has led
the United States Federal Administra-
tion to establish a small, high level,
expert group to develop an alternative
land use/transport modelling process.
This is the most significant project
in hand in the area of city transport
planning, and is intended to yield a
process for rapid assessment of the
consequences of proposed transport
systems in urban structure alternatives.”

4. MAJOR ASSUMPTIONS UNDERLYING THE MODEL

4.1 Introduction

The future level of traffic will depend upon a
number of variables. The most important are:-

The future population and its
location
The future level of employment
and its location
The future of certain other
important traffic generators
(the port, the airport etc.)
The growth in car ownership
The price and the availability of
fuel
The growth in the economy and the
level of employment and unemployment
(the workforce participation ratio)

In the application of the traffic model the State
Transport Study Group is required to apply its
corporate mind to each of these matters. Predic-
tions must be made for the forecast year 15 years
ahead.

We will, in this section, briefly examine the
methodology employed.

4.2 Estimates of Population and its Location

Population is a fundamental ingredient of the model.
Errors in population estimates will manifest them-
selves throughout the various sub-models (the number
of trips generated or attracted; the level of car
ownership and so on).

-163-

We have, in the course of this exposition, made
reference to the difficulty of accurately forecasting
population. We have, somewhat unkindly, drawn
attention to errors made in the past. The Cumberland
Plan substantially underestimated population.

The Sydney Region Outline Plan made an overestimate
of population.

There are two specific problems. First, it is
difficult to calculate the rate at which the population
will grow. Secondly, it is difficult to say where
that growth will be manifested. Yet traffic forecasts
will be wide of the market unless population estimates
are accurate in both respects.

Population growth is the product of three things:-

The mortality rate
The fertility rate
The net gain ( if any) as a
result of immigration (i.e., the
number of immigrants arriving
less the number of residents
(including immigrants) leaving)

There has been a marked decline in fertility. It is
a phenomenon experienced by most Western countries.
A number of explanations have been suggested. First,
there has been a greater participation of women in
the workforce. The percentage of women over the age
of 15 working has risen dramatically. The statistics
are ₍₄₆₎:

In 1947, 24.9% of women under
the age of 15 worked
In February 1973 the estimated
percentage was 40.5%

46. Summary of Population and Planning: “The
    Consequences of Ignoring Demographic Reality”
    Robert Birrell, Volume 13, Royal Australian
    Planning Institute Journal, page 88.

-164-

Secondly, there have been improvements in contracep-
tion. Thirdly, there has been a fundamental shift
in attitude on matters important to fertility.
People are marrying later or not marrying at all.
They are having their children later, and they are
having less children.

The dramatic nature of the shift is underlined by
the following table ₍₄₇₎:

TABLE 14.

NUMBER OF CHILDREN CONSIDERED IDEAL
IN AUSTRALIA (PERCENTAGE OF RESPONDENTS)

Number of children	1947	1958	1973	1974
Two or less	15	18	40	47
Three	21	25	28	28
Four or more	62	50	27	23
Unsure	2	7	5	2

Immigration, to an even greater degree, is subject
to fluctuation, because it is dependent upon policy.
Policy can change overnight. There are also
difficulties created by interstate and intrastate
immigration. These movements are difficult to
monitor. It is, therefore, especially difficult to
accurately discern a trend.

For all these reasons it is hardly surprising that
population projections are quickly superseded. The
New South Wales Planning and Environment Commission
in its projection of August, 1976 said this (48):

47. Birre11 ibid., page 88.
48. Technical Bulletin No 8, Planning and Environment
Commission of New South Wales, page 9.

-165-

“Population projections become out
of date quickly and must be revised
regularly. Some of the assumptions
made in Technical Bulletin No. 4
(November, 1973) now need to be
revised in the light of marked changes
in demographic trends since 1971."

The Leitch Committee said much the same thing ₍₄₉₎:

"Our attention has been drawn in
evidence to the difficulty of forecas-
ting population and the uncertainty
which must necessarily arise over the
future estimates of this factor. We
give (in a graph)..various projections
of population which have been made at
different points in time in the recent
past, and the variation one from
another is clearly considerable, which
illustrates not only the difficulty of
arriving at usable population projec-
tions but also the rapidity with which
projections can change."

(emphasis added)

Bruton makes the following assertion in respect of
population estimates ₍₅₀₎:

"Changes in the technological, and
demographic spheres ensure that it
is impossible to predict accurately
population trends for more than 5-10
years ahead..."

The Sydney Region Outline Plan made a similar
comment ₍₅₁₎:

"Demographic projections beyond 10
years are notoriously uncertain,
especially where substantial migration
movements are involved."

The estimates made by the transport model are based
upon a 15 year projection.

49. Report of the Advisory Committee on Trunk Road
Assessment (Leitch Committee), page 83.
50. Bruton op.cit page 75.
51. Sydney Region Outline Plan, March, 1968, page 18.

-166-

Even more formidable than the task of accurately
projecting the population growth, is the task of
successfully identifying where that growth will
occur. The following words of caution accompany
the estimates given for local government areas in
August, 1976 ₍₅₂₎:

"A demographic projection at this leve1
(i.e., local government area level) is
not feasible...as there are too many
variables which may cause quite erratic
and sudden alterations in trends. For
example re-zoning of land by local Councils
can change an area's permitted population
density, while delays or accelerations
in the phasing of land releases can alter
the timing of population growth. . . .
Consequently the figures for local
government areas will be ‘estimates'
rather than 'projections’."

The Inquiry had the benefit of certain views expressed
by Mrs. Carol Ivison, a demographer with the Planning
and Environment Commission. She said ₍₅₃₎:

"The smaller the area, therefore, the
greater the problems (of estimating
population) and the greater the effect
an error has. For local government areas
the problems are many and some appear to
be insoluble. At this 1eve1, accurate
demographic projections are not possible
because constraints other than demographic
influences are of paramount importance.
Availability of land, which is not an
issue at State or Regional level, becomes
crucial. . other variables also may cause
sudden or erratic alterations to trends.
For example, changes in zoning or Council
policy may encourage or impede development.”

Such changes can have important consequences for
traffic (and, therefore, for traffic estimates).
The Sydney Area Transportation Study furnishes an
illustration. It will be recalled that it adopted
the population estimates of the Sydney Region Outline
Plan (1968). Technical Bulletin No. 4 (November, 1973)

52. Technical Bulletin No. 8, Planning and Environment
Commission of New South Wales, page 31.
53. S.K/C 947 Submission of the Planning & Environment
Commission of New South Wales, Appendix 1, page 12.

-167 -

and its addendum reflected certain changes in trend.
The changes were not simply changes in the total
population; rather they were changes in location.
They are described in the following passage ₍₅₄₎:

"However, the more compact and denser
land use pattern of TB4 allows people
to live closer to jobs, schools and
other facilities, causing significant
increases in the proportion of short
intrazonal trips. These short trips
would use loca1 streets and local bus
services rather than regional freeways,
expressways and railways..."

The traffic model is based upon the population fore-
casts made by the Planning and Environment Commission
in August, 1976. Shrewdly, it adopted the low forecasts
from amongst the range of forecasts suggested at that
time. The most recent publication by the Commission ₍₅₅₎
indicates a further diminution in the rate of
growth. The difference between the 1976 'low’ forecast
and the 1979 ‘medium’ estimate is revealed by the
following table ₍₅₆₎:

The Table reveals that the Study Area under scrutiny
by this Inquiry is more or less stable, and expected
to remain so. The important issue, from the viewpoint
of traffic, is where the growth will occur. The
Study Group put the matter in this way ₍₅₇₎:

"LEAVENS: The inner areas which are fully
built show remarkable stability over time.
The population may vary 10%. By the mere
fact that all of the urban land is utilised,
they don't change their population very
much. (The growth) tends to be in the
outer areas, the fringe areas, which
generally accommodate...the ultimate
growth in the region, which show the
remarkable variations."

54. UTSG a revision of SATS Year 2000 Travel Forecasts,
    July, 1975, page 10.
55. Population Projection for N.S.W. 1976-2001, PEC,
    February, 1979.
56. The table summarises Table 1 forming part of
    Appendix 1 to the submission by the PEC to the
    Inquiry S.K/C 947.
57. Transcript UTSG (now STSG) 23.11.79, page 35.

-168-

TABLE 15.

COMPARISON OF THE LOW ESTIMATE TECHNICAL BULLETIN NO. 8
AND THE MEDIUM ESTIMATE, FEBRUARY, 1979

-169-

In a discussion paper circulated by the New South
Wales Planning and Environment Commission ₍₅₈₎,
the problems of identifying where increases in
population will be located are underlined. Three
possible scenarios are suggested. Each will have
different traffic implications. Which (if any) is
closer to the truth is anyone's guess. The first
possibility is a continuation, more or less, of
existing trends. The built-up area will continue
to lose population to the outer areas, though at a
reduced rate. The distribution suggested by this
scenario is as follows ₍₅₈₎:

                            1976-2001
Alternative 1     Population increase

Built up area               - 100,000
Western sector                200,000
South-west (Macarthur)        230,000
North (Gosford/Wyong)         200,000
Scattered                      50,000
                            ---------

Sydney Region 580,000

The second alternative presupposes a change in
direction. There are straws in the wind which do
suggest such a change. The change is towards urban
consolidation. The Planning and Environment Commission
put it this way ₍₅₉₎:

"However it, is possible that (the)
situation will alter. While there is
nothing in past trends to indicate any
immediate change in the next 20 years,
there are likely to be new factors which
might affect population distribution.
These include smaller family size, greater
married female participation in the work-
force and the foreshadowed energy crisis,
one result of which could be an increased
reliance on public transport. The
combined effect of such factors could
result in less emphasis on the Australian
traditional preference for large detached
blocks of land, and increase the trend
towards medium density houses which are

58. Sydney Region Population Estimates for Local
Government Areas 1976-2001 – Discussion Paper,
September, 1979 (79-11). Page 8.
59. Discussion Paper ibid., page 8.

-170-

similar in style but smaller in scale.
An increase in the number of childless
couples may also result in additional
flat building in some areas."

The paper then refers to certain matters which
signify a change in direction. It said ₍₆₀₎:

"There have also been Government
decisions recently directed towards
revitalizing the built-up area,
including policies supporting urban
consolidation and dual occupancy."

A plausible distribution based upon this view is as
follows ₍₆₀₎:

                            1976-2001
Alternative 2     Population increase

Consolidation in
   built up area              140,000
Western sector                250,000
South-west (Macarthur)        117,000
North (Gosford/Wyong)          73,000
                            ---------

Sydney Region 580,000

A third alternative is suggested. It presupposes
‘strong support (by the Government) for the growth
centres’. The distribution under this alternative
is as follows ₍₆₁₎:

                            1976-2001
Alternative 3     Population increase

Consolidation in
   built up area               70,000
Western sector                160,000
South-west (Macarthur)        150,000
North (Gosford/Wyong)         200,000
                            ---------

Sydney Region                 580,000

60. Discussion Paper ibid., page 8.
61. Discussion Paper ibid., page 9.

-171-

The variations may be less important for some of
the road options than for others. Within the area
primarily served by the Cooks River option the
population is remarkably stable. The South Western
Option (at least to the west of Ring Road 3)
depends, in large measure, upon growth in the
Campbelltown Region (the Macarthur Growth Centre).
In the evaluation of options we will examine the
evidence on that issue.

4.3 Employment and Land Use

It will be remembered from our discussion of the
model that population estimates are crucial to the
trip generation equations. They te1l you how many
people each day will set out for work from a
particular zone. It is necessary to then predict
where those people will be travelling. This will
depend upon where jobs are located and the way in
which land is used. How is this done? The Study
Group gave the following evidence ₍₆₂₎:

"LEAVENS:..for land use information we
rely almost entirely upon the SATS work.

COMMISSIONER: And how is it that they
arrive at their material? Did they do
it simply by local government planning
schemes, by the Cumberland Scheme or what?

LEAVENS: ..an ordinance made it mandatory
for a land use map to be prepared for
each local government area prior to the
prescription of a scheme and I was told
that this formed the basis of most of the
local government area land use information
collated for SATS in 1971."

The problem is not greatly different from estimating
population. There is the 'sunk-in' investment - the
established industrial and retail areas - about which
predictions can be made with some confidence. The
greatest uncertainty must be the extent to which
industry will absorb vacant industrial land on the
fringes or within existing industrial estates. One can
establish trends and growth rates.. As with population.

62. Transcript UTSG (now STSG) 4/12/79, page 15.

-171-

they have the habit of changing. Government
initiatives (including the building of a road) can
influence that change.

Even the capacity of an existing industrial area
to generate traffic can change, and change rapidly.
The Study Group in one of its publications said
(referring to the Central Industrial Area) ₍₆₃₎:

"..Because it is difficult to accurately
predict future travel, we will also
carry out sensitivity tests to test the
effects of high and low trip generation
levels in different parts of the (Central
Industrial Area) ...even within the existing
zoning of the area, land uses are changing
rapidly."

The Planning and Environment Commission in its
submission to the Inquiry referred to some of the
changes in the Central Industrial Area, and their
traffic implications ₍₆₄₎:

"There has been a noticeable increase in
the number of distribution and warehousing
facilities built in the vicinity of the
Port (especially Botany) and it appears
these are replacing older manufacturing
buildings. Conversions of obsolete buil-
dings in the South Sydney and Botany areas
to accommodate container traffic has been
quite pronounced, and this trend is
expected to intensify..These developments
are land intensive, have much fewer labour
requirements than manufacturing (which it
is replacing) and generate increased
industrial traffic which may severely
affect loca1 community environmental and
residential amenity standards."

Quite apart from these issues, the Study Group
required to apply its mind to the future (in 15 years)
of the Botany Bay Port and Airport. They go to
enormous lengths to ferret out the most recent
information. In the case of the Airport they had
access to the aviation forecasts prepared by the
MANS Study.

63. UTSG (now STSG) Central Industrial Area Study
Discussion Paper No. 2, July, 1976, page 16.
64. Submission PEC S.K/C 947, Appendix 2, page 1.

-173-

However, it is not simply a matter of diligence or
even perspicacity. There are matters associated
with the future of Mascot Airport and Port Botany
which are essentially unforeseeable. The outcome
of this Inquiry and its recommendations (and the
Governments reaction to those recommendations) is
but one. This Inquiry has recommended the use of
rail for containers destined for the western suburbs.
That recommendation, if it finds acceptance with the
Government, will affect traffic. Its affect will
extend beyond the number of container vehicles
involved. The container vehicles have a dispropor-
tionate impact upon the public mind and upon the
environment. Their accessibility needs are far
greater than commuter traffic. Though they may be
only one or two percent of the traffic stream, they
are the percent that matters. Policy consideration
of this sort inevitably affect many land use decisions.
In consequence they must affect the future flow of
traffic.

4.4 The Growth in Car Ownership

4.4.1 The Approach by the SATS Study

The SATS Study boldly stated its assumption when
calculating growth in motor vehicle ownership ₍₆₅₎:

"It was assumed that the motor car, in
its present general form, would continue
as the main mode of personal transporta-
tion until at least the end of the
century - and that adequate fuel, of
some kind, would be available throughout
the entire period."

(emphasis in the original)

The Study Group model makes the same assumption.

By an examination of the growth in motor vehicle
ownership in the past, and by employing the
statistical technique, regression analysis, a
forecasting 'model' (i.e., a mathematical formula)

65. SATS Volume 2, Chapter III, page 36.

-174-

is constructed. The formula employs the logistic
curve. This is an ‘S' shaped curve. The shape of
the curve is shown in Figure 8. It presupposes
the determination of a saturation level. A point
must be reached, so the argument runs, where everyone
eligible to own a car (those over the age of 17 years)
in fact owns one. In the United Kingdom the upper
limit of the saturation level is calculated by
making the following assumptions ₍₆₆₎:

"..an upper limit for the saturation level
has been obtained by assuming that 90%
of the population aged 17-64 and 50%
of those aged 65 and over will eventually
own a car. This gives a saturation level
of 0.6 using the 1972 age distribution.”

The Transport and Road Research Laboratory, however,
showed a lower saturation level for the purposes of
its calculations. They concluded ₍₆₇₎:

"None of the methods described above
gives by itself a convincing estimate
of saturation level, and the estimate
vary somewhat.. In previous forecasts
a saturation level of 0.45 has been
used. It is believed that the data
presented may justify a higher figure,
but until the interpretation of the
National Time Series Data is clearer
the previous figure will be retained."

The SATS Study adopted a higher figure. It said ₍₆₈₎:

"Based on New South Wales and California
trends, the most likely saturation
leve1 is 0.55 vehicles per person."

The saturation level is determined independently of
the effects of income and motoring costs. The growth
in car ownership will be highly sensitive to the
saturation level chosen ₍₆₉₎.

66. Leitch Committee ibid., page 11.
67. Leitch Committee ibid., page 11.
68. SATS Volume 2, Chapter III, page 36.
69. Leitch Committee ibid., page 86 and Transcript
UTSG (now STSG) 4/12/80, page 66.

-175-

FIGURE 8.

CAR OWNERSHIP AGAINST TIME
(THE LOGISTIC CURVE USED IN THE UNITED KINGDOM)

SOURCE: Report of the Advisory Committee on Trunk Road
Assessment (The Leitch Committee Report), page 9.

-176-

4.4.2 Criticism of the Car Ownership Model

The Leitch Committee in the United Kingdom was not
convinced that a car ownership model, essentially
similar to that employed in the Sydney traffic model,
accurately portrayed the likely growth in car owner-
ship. They offered a number of criticisms.
Reference has been made to one already, though it
bears repeating. The model implicitly assumes that
the growth in car ownership is a good thing, and
should not be discouraged by any policy on the part
of the Government. The Leitch committee said this (70):

"Commentators believe variously that the
Government should decide whether the
levels of car ownerships implied by the
forecasts for the future are desirable
and, if not, should take steps to reduce
them; that a policy of encouragement of
other modes should be pursued to slow
down the rate of increase of car owner-
ship implied by the forecast..”

Curiously, the SATS study did confront this policy
question. It did not mince words. It said ₍₇₁₎:

"Suggestions have been made, from time
to time, that restriction should be
placed on motor vehicle ownership.
This is difficult to achieve in a
democracy - a better solution is the
construction of high speed, limited
access roads to provide adequate
mobility to the large number of private
and commercial vehicle movements."

Since that time the mood has changed. The winds of
change were already blowing in 1973 when the Commonwealth
Bureau of Roads wrote the following (72):

"The dominance of road travel, especially
by car, has caused many observers to
question whether the nation can afford
what they consider to be the high-cost
of car travel."

70. Leitch Committee ibid., page 51.
7L. SATS Volume 2, Chapter I, page 5.
72. Report on Roads in Australia, Commonwealth Bureau
of Roads, 1973, page 54.

-177-

Since 1973 the mood has changed still further.
Uncertainty about energy, disenchantment with
freeways within the inner-city area, the renewed
interest in public transport, and policies aimed
at minimising the travel needs of the city, are all
symptoms of that change.

Secondly, the use of the logistic curve is criticised.
The Leitch Committee said ₍₇₃₎:

"The question then arises how best to fit
a curve to the known data and what
trajectory it should follow into the
saturation level. The logistic curve
is only one of the many forms of curve
that it is possible to fit to the known
data. Thus there is nothing uniquely
right about the ‘S’ shaped curve that
determines the level of car ownership
in future years.
(emphasis added)

Finally, the Leitch Committee is critical of the
concept of a ‘saturation level’. The saturation
level is of the utmost importance in determining
the rate of growth. Yet it is not possible to
arrive at any satisfying estimate of the level of
car ownership which might be considered ‘the
saturation'.

Ultimately the Committee advocated that the Department
of Transport (responsible for modelling) should give
consideration to a 'causal model' of future car
ownership. It said ₍₇₄₎:

"Those who advocate the use of causal
models argue that the decision to own
and use a car is a complicated one, which
can only be predicted by a full under-
standing of the relevant factors..The
difference between the two methods can
be easily stated: Extrapolatory methods
make 1ittle attempt to represent expli-
citly the causal factors that underlie
car ownership and use while causal ones
do. We believe that better forecasts can
be obtained if what are taken as the
underlying causal factors are introduced
explicitly than if they are not."

73. Leitch Committee Report ibid., page 87.
74. Leitch Committee Report ibid., page 88.

-178-

4.4.3 The Performance of the Car Ownership Model

Yet notwithstanding all criticism, it appears that
the car ownership model has performed very well thus
far. The Study Group’s own assessment appears from
the following passage ₍₇₅₎:

"FIELD: Now that equation (the car
ownership equation) is a trend equa-
tion, the only variable being time.
It assumes that there is some satur-
ation level of car ownership which
we are progressively heading towards..
the particular technique is highly
open to question; it has absolutely
no input other than time. It does
assume, for instance, that basically
car buying and ownership and real
income habits will continue in the
future as they have in the past.

Up until 1976 that equation has been
spot-on – virtually exact for the ten
years since the equation was calibrated.
There is possibly a good argument in
the future why we should down-grade
that. It was not done. The straight
trend was used for this study.

LEAVENS: Perhaps I could give the
figures..In 1971 (the SATS Base Year),
using the equation you get a figure of
0.3125 vehicles per person. In 1976
using the equation again you get a
figure of 0.3678 vehicles per person.
The 1976 Census reported a factor of
0.367 vehicles per person so the
equation is holding to the fourth
decimal place."

(emphasis added)

Yet, as the Study Group suggests, we may be on the
threshold of change. A Bulletin issued by the
Australian Bureau of Statistics in September, 1979,
for instance, indicated a fall in the registration
of new motor vehicles compared to the previous year.
On an annual basis it appears that motor vehicle
registrations were increasing (as they have in the
past) though at a slower rate. There are grounds for
believing, therefore, that the motor vehicle ownership
figures derived from past trends may exaggerate (to
some degree) the growth of motor vehicles in the future.

75. Transcript UTSG (now STSG) 4/12/79, pages 63-64.
76. Australian Bureau of Statistics Bulletin, 24/9/79.

-179-

4.5 Economic Growth and Workforce Participation

The model assumes stability. It assumes the economy
will continue to grow as it has in the past. The
economy, as such, is not explicitly identified in
the variables measured and predicted. Indirectly
it is reflected in almost every variab1e. This
comes about because these variables are projected
into the future based upon past trends. The past
trends were established during a time of economic
growth.

In the case of workforce participation the assumption
is ₍₇₇₎:

"A workforce participation ratio of
0.42 (about equal to 1976 estimate).

If there were a sudden increase in unemployment as a
result of changes in the economy, or the introduction
of technology, or for whatever reason, the workforce
participation ratio (the ratio of people working to
those not working) would alter. Traffic would alter
in consequence.

At this moment in time it seems reasonable to
suppose that the economy will continue to grow.
Events essentially outside the control of Australia,
however, may alter that prospect.

One does not have to presuppose matters of a gloomy
or catastrophic nature to upset the essential
stability of past trends. It is not inconceivable
that the practice, evident in some industries, of
people working a four-day week and enjoying three
days of leisure (and perhaps working shorter hours)
may become more wide-spread. This may have signifi-
cant effects upon traffic patterns. It is but one
of the many uncertainties which lie in the future.
Only the future can reveal when, if ever, it comes
to pass.

77. DMR Submission S.K/C 340, Transport and Economic
Analysis, page 3.

-180-

4.6 The Price and Availability of Fuel

4.6.1 The Issues Which Must be Addressed

There are three issues:

First, is the judgement made by
the Study Group, that the price
of fuel will double in real terms
by 1991, reasonable?
Secondly, is it reasonable to
suppose that the supply of petrol
will not be disrupted?
Thirdly, what effect will a price
increase (or a disruption in supply)
have upon travel habits?

The issues are intertwined. It is necessary to
differentiate between Australia's position as a
petrol consumer and the world position. To what
extent is Australia likely to be dependent upon
imported oil rather than locally produced oil?

We will begin by comparing the assumptions made by
other Inquiries where energy has been an issue. We
will briefly refer to some of the matters (other
than supply) which may effect price. We will then
deal with the issue of supply.

Libraries have been written on the ‘energy crisis’
and its likely effect. It is beyond the scope of
this Report to do more than distil some of the
documented facts and the better rehearsed theories.
It is an issue which excited the attention and
imagination of the many people who made submissions
to the Inquiry. It is therefore important that it
should be addressed and a judgement made.

-181-

4.6.2 The Judgement of Other Inquiries

The Advisory Committee on Trunk Road Assessment
(the Leitch committee) in March 1977 received
advice from the Department of Energy in the United
Kingdom to this effect ₍₇₈₎:

"Oil prices are likely to continue
to rise in the longer term and may
well double in real terms by the
turn of the century."

The Study Group postulates a doubling (in real
terms) by 1991.

The committee enquiring into the airport needs of
Sydney (the MANS Study) took a supremely optimis-
tic view when making its aviation forecasts in
December, 1977. It assumed ₍₇₉₎:

“World crude oil prices will show
zero growth in real terms (after
allowing for inflation)
Domestic crude oil prices will be
increased to import parity by 1985
(to equal world price)
At these prices, sufficient fuel
will continue to be available over
the forecast period.”

Such are the hazards of forecasting that import
parity was introduced less than a year after the
making of this statement (August, 1978).

The year 1973 marked a watershed in petrol pricing.
OPEC (Organisation of the Petroleum Exporting
Countries) determined upon a new pricing policy.
Prices were fixed according to what the market could
bear. The Petroleum Gazette describes the change in
policy, and its implications, in these words ₍₈₀₎:

78. Report of the Advisory Committee on Trunk Road
Assessment, page 204.
79. MANS Study information bulletin No 1. ‘Aviation
Forecasts’, page 5.
80. Petroleum Gazette, September, 1979, page 65.

-182-

"OPEC then determined (16.10.73) that
hence forth the pricing of petroleum,
like the pricing of other internationally-
traded manufactured goods, commodities
and raw materials, should be market-
oriented. Results were swift.

The October 1973 increase was 70%, with
the staggering rise of 30% at the end
of the year.”

The reaction overseas was immediate and severe. It
was made the more severe in the United States by an
embargo imposed by a separate, but kindred, organi-
sation known as OAPEC (Organisation of the Arab
Petroleum Exporting Countries). As a response to
the Arab-Israeli war which broke out on the 6^th
October, 1973, OAPEC imposed an embargo on the United
States dating from the 17th October, 1973.

Australia was insulated from this turmoil. It was
then, and is now, substantially self-sufficient in
oil (65-70%). That may change. We will examine
the evidence shortly. The change in pricing in
Australia did not occur until the Federal Government
adopted, in August 1978, a policy of full import
parity. There had, incidentally, been a policy of
partial parity pricing dating from September 1975.

The movement in prices compared to average weekly
earnings is depicted in figure 10. The figure is
extracted from the Economic and Transport Analysis
of the Department of Main Roads ₍₈₁₎.

The graph is accompanied by the following statement ₍₈₂₎:

"It can be observed (from figure 10) that
after a steady decline over a period of
30 years, the price of petrol to the
consumer has risen sharply in the past
year. The recent rises have been due
partly to the rises in the price paid
for overseas oil and partly due to
governmental decision to artificially

81. DMR Submission S.K/C 340, page 4.
82. Submission ibid., page 5.

-183-

FIGURE 9.

COMPARISON OF RETAIL PETROL PRICES WITH
AVERAGE EARNINGS (in 1978 dollars)

Sources: 1. Australian Bureau of Statistics

2. Australian Institute of Petroleum Limited

-184-

increase the price of locally produced
oil. Had the governmental decision been
taken at the same time as the price for
overseas oil started to rise, the slope
of the petrol price curve at the right
hand end would have been flatter. The
interesting point from figure (10) is,
however, the fact that the current price
of petrol, in real terms, has not yet
reached the prices charged prior to
1952."

Commenting upon the model the Department then said ₍₈₃₎:

"It is unfortunate that the travel
forecasting model used in this study
is not sensitive to the effects of
variations in the (price and availa-
bility of fuel) on the total amount
of travel. On the other hand, it is
considered that the values assumed
for these factors can be placed in
the 'very likely' categories."

The assumption used in the model (that the price of
the petrol would double in real terms by 1991) was
adopted in 1977. The events in Iran began in
December 1978 with the overthrow of the Shah. They
could not have been foreseen in 1977. In the light
of those events the Study Group was asked whether
it wished to revise its assessment of the likely
future price of petrol. The question and answer
were as follows ₍₈₄₎:

"COMMISSIONER: With hindsight are you
prepared to concede that the assumption
which lies behind the forecasts, namely
that the 1991 cost of petrol will be
double in real terms the 1971 costs,
was in fact probably in error?

FIELD: No, not at all. We would tend
to think that is still a reasonable
estimate."

In July 1978 the Australian Transport Advisory
Council prepared a ‘Transport and Energy Overview’.
On the question of pricing the document had this
to say ₍₈₅₎:

83. ibid., page 6.
84. Transcript UTSG (now STSG) 4/12/79, page 95.
85. Transport and Energy Overview, page 1.

-185-

"It is evident there are considerable
margins for error in estimating the
rate at which the world and Australian
petroleum prices will rise..”

The Study Group, no doubt, would not disagree.
Doing the best it can it has estimated a doubling
of the price. The Inquiry, for its part, has no
better information upon the basis of which it can
dispute that judgement. It has the uneasy feeling,
as does the public, that the situation may get
somewhat worse than the model postulates.

However, it should be said that the price of crude oi1
is not the only factor affecting the retail price
of petrol. Indeed, curiously, it is not even the
most important factor, though its influence is
considerable ₍₉₆₎:

"The impact of the rise and the cost
of oil to refineries of the order of
90% will be much less dramatic on the
transport consumer since the cost of
crude is currently less than 2O% of
the selling price of refined fuels,
the balance being taxes, refinery
costs, marketing, distribution and
retailing costs."

We have, thus far, artificial1y separated the issues
of price and supply. Obviously the one is related
to the other. We will now turn to the question of
supply. We will first examine the supply of oil
available within Australia. We will then examine
the world prospects for this precious commodity.

4.6.3 The Supply of Oil in Australia

Australia's position will depend upon the extent to
which it can satisfy its own needs. It is, therefore,
necessary to address the following:

86. Transport and Energy Overview, page 70.

-186-

the known oil stocks within Australia
the comparison between supply and the
projected demand
the prospect of new discoveries
the prospect of substitutes being
found to take the place of oil

We will examine these questions in turn. We should
first, however, draw attention to the energy demands
within various sectors, and specifically, the
contribution made by cars. In a publication by Esso
entitled 'Australian Energy Outlook' (October 1979)
there is a graph which is worthy of reproduction ₍₈₇₎:

FIGURE 10.

AUSTRALIAN ENERGY DEMAND

The contribution by cars is 57% ₍₈₈₎ of the transport
demand.

The Esso publication reproduces a diagram which is
now familiar to most Australians through television.
The diagram appears below ₍₈₉₎:

87. Australian Energy Outlook, Esso, page 8.
88. Article 'Energy: Where are We Heading?’ In
Truck and Bus Transportation Journal, December
1979, page 36.
89. Australian Energy Outlook, ibid., page 14.

-187-

FIGURE 11.

AUSTRALIA - OIL SUPPLY/DEMAND (NO
NEW OIL DISCOVERIES)

The conclusion drawn by Esso is in the following
terms ₍₉₀₎:

"Stated simply, if there are no new
discoveries, Australia will be
importing 45% of its oil supply by
1985 and about 65% by 1990.
In less than a decade, Australia
could move from being 65% self-
sufficient in energy from oil to a
nation dependent on increasingly
uncertain international sources for
the majority of its supply.”

4.6.4 The Prospects of New Discoveries

In another publication by Esso 'Oil and Gas - The
Production Story' the following gloomy prospect is
offered ₍₉₁₎:

"There is little doubt that most of
the 'easy oil’ in the world has now
been found and that to discover and
produce the remaining deposits will
be progressively more difficult and
costly."

90. Australian Energy Outlook ibid., page 14.
91. Oil and Gas – The Production Story, page 7.

-188-

There are known reserves which are considered
either technologically or economically out of reach.
The greater the rise in fuel prices, the greater the
likelihood of passing the cost threshold. The North
Sea oil exploited by the United Kingdom is one
example. There have also been (and one can expect
there will continue to be) advances in technology,
facilitating the recovery of oil reserves.

What are Australia's prospects? The Transport and
Energy Overview offers the following pessimistic
prognosis ₍₉₂₎:

"With regard to Australia’s potential
reserves there appears to be a consensus
that Australia's prospects of having
large reserves of oil are not good.
It is also generally thought that what
oil may exist is probably in offshore
high cost regions and hence the Australian
explorer must be prepared to pay more
per barrel to produce oil than in many
other areas of the world."

Later in the same report a glimmer of hope is
contained within the following ₍₉₃₎:

"Also in recent times there appears
to be some doubt being cast upon the
conventional wisdom that Australia’s
oil prospects are not good..Although
it is not thought that large deposits
will be found onshore, the lower
production costs together with the
world parity price would make many
smaller onshore deposits commercial.
Australia is thought to have extensive
marginally hopeful onshore areas which
may contain commercial deposits under
these circumstances."

In the Esso publication ‘Australian Energy Outlook’
the oil supply is compared to demand, taking account
of new discoveries (those which are considered to be
‘medium probable’). It can be depicted as follows (94):

92. Transport and Energy Overview, page 48.
93. Ibid., page 50.
94. Australian Energy Outlook, Esso, page 18.

-189-

FIGURE 12.

AUSTRALIA - OIL SUPPLY (WITH NEW DISCOVERIES)

Even where new discoveries are made, there is a
long lead time (estimated variously at between 4
and 10 years) between the time of the discovery
and the time the crude oil can be utilised. Writing
in 1978 the Transport and Energy Overview had this
to say ₍₉₅₎:

"However, even if large discoveries
were made today it is unlikely that
they could come onstream before 1985.
Therefore, it seems that Australia
cannot avoid increasing import
dependence in the short term.”

When that was written the Bass Strait Fields were
firmly in decline ₍₉₆₎. The period before the Bass
Strait Fields begin to sharply decline seems, from
more recent assessments, to have been pushed back
until approximately l988.

95. Transport and Energy Overview ibid., page 63.
96. See Canterbury Council Submission S.K/C 341,
    page 12 and (also the Total Environment and
    Friends of the Earth S.K/C 1287) to the same
    effect.

-190-

4.6.5 The Prospects of a Petroleum Substitute

The search for alternative sources of energy is a
subject as fascinating as it is vast. In the context
of transport the possibilities can be divided into
two streams:

First, the development of new
liquid fuels to take the place of
petrol. They include synthetic
petrol derived from coal (such as
that being produced by the South
African Government at Sasol) or the
use of L.P.G. or Alcohol derivatives.
Secondly, the development of a new
energy source requiring major changes
to existing transport technology.
The electric car is perhaps the best
example.

In all cases long lead times are contemplated. The
Transport and Energy Overview concludes with these
remarks ₍₉₇₎:

"The close matching of petroleum fuel
characteristics in road transport
technology will preclude any widespread
use of petroleum fuel substitutes in
road vehicles in the short term future.
L.P.G. could, however, find immediate
application in certain specialist
vehicles resulting in small but signify-
cant petrol savings. In the longer term
an alternative to fossil oil derivatives
fuel must be found but the production
technologies of most alternative fuels
are only nascent and their economics
uncertain. Synthetic petrol derived
from coal is, at the moment, one of
the strongest candidates for meeting
our long term transport fuel requirements
while the use of methanol in admixture
with petrol could significantly reduce
transport petrol consumption in the
nearer future. Both these options,
however, require large capital invest-
ment in production planning.”

97. ibid., page 166.

-191-

4.6.6 What, then, Are the Prospects for Australia?

The major options under contemplation will not be
completed before 1991, even on the most optimistic
assessment. It is likely that they will be completed
well after that date, if they are built at all. The
Inquiry must, on the energy issue, concern itself
with a period extending beyond 1991 (being the date
selected in the transport models). Having regard to
the foregoing, what are the prospects?

The Transport and Energy Overview paints a picture
which is less than enticing ₍₉₈₎:

"Australia's reserves are relatively low
compared to its rate of consumption, and
the probability of finding large new oil
deposits is thought to be low. The
prospect is one of rapidly increasing
dependence on OPEC oil, large oil price
increases and insecurity of oil supply. "

The Petroleum Gazette for December 1979 said this ₍₉₉₎:

"So although Australia’s known recoverable
oil reserves total 2.1 billion barrels,
in the next 10 years this country's
expected petrothirst will demand a total
of 3.4 billion barrels.

That is 50% more than all of the known
available domestic oil supply, with all
fields effectively pumped dry and
nothing left over for the future. Of
course, the fields are not being produced
as hard as that; but they are being
depleted quite rapidly."

Reduced to percentages they say this ₍₁₀₀₎:

"If there are no new discoveries of oil,
Australia (now 65% self-sufficient in
oil energy) will be importing 45% of
its oil supply by 1985 and 65% by 1990."

The gap between supply and demand is likely to effect
transport rather more than other sectors. The Bass

98. Transport and Energy Overview, page 1.
99. Petroleum Gazette December 1979, page 76.
100. ibid., page 77.

-192-

Strait oil is especially suitable for use by
transport, and is substantially used for that
purpose at the present time. The prospects of
transport sector having to import from overseas are
therefore, rather greater than other sectors ₍₁₎.

4.6.7 The World's Supply of Oil

We are now conscious (we once were not) that there
is a limit to the fossil fuels available on Earth.
There is a need to conserve energy and find alterna-
tives. It is artificial, however, to postulate the
reserves suddenly drying up ₍₂₎. The threat to the
world's oil supply is rather to be found in the
volatile and capricious nature of the market.

To understand the vulnerability of Australia, if
and when it has to rely upon that market, it is
necessary to examine the following:

Who controls the market?
From whom does Australia import
petroleum?
The nature of OPEC?
The 1ikely future supply of oi1?

These issues will be examined one by one.

4.6.8 Sources of Australian Imported Oil

The world market for the ‘non-communist world' is
firmly in the control of the OPEC nations. Their
dominance can be appreciated from the following
diagram ₍₃₎:

1. Transport and Energy Overview supra, page 37.
2. Report of the Advisory Committee on Trunk Road
Assessment (the Leitch Committee), Appendix F,
page 203.
3. Petroleum Gazette December 1979, page 79.

-193-

FIGURE 13.

DISTRIBUTION OF KNOWN OIL RESERVES

The dependence of Australia upon the OPEC countries
(and Saudi Arabia in particular) emerges from the
following table ₍₄₎:

TABLE 16.

SOURCES OF AUSTRALIA'S IMPORTED OIL 1975-76

4. Extracted from Transport and Energy Overview,
page 55.

-194-

Overall 82.8% of Australia's import of petroleum
comes from the Middle East.

4.6.9 The OPEC Organisation

The Organisation of the Petroleum Exporting Countries
(OPEC) was founded in Baghdad in September 1960. It
started with five members of which Venezuela was the
only non-Arab state. The founding members included
Iraq, Saudi Arabia, Iran and Kuwait.

The organisation has expanded to fifteen nations. It
includes Indonesia, Libya, Nigeria and Equador.

We have already referred to the events of 1973. The
escalation in the price of oil since that time is a
matter of such notoriety that it need not be laboured
here.

What should be repeated, and what is less widely
appreciated, is the threat to supply. It is unlikely
that the tap will be turned off completely. The OPEC
countries depend upon foreign exchange for their
imports. They have invested heavily in the West.
They are concerned with the continued vitality of
Western economies to sustain their investments.

Rather, the prospect is of a diminishing supply in a
world habituated to economic growth, and hence, an
accelerating demand. Libya in 1970 restricted its
production. In October 1973 OPEC announced a reduc-
tion of 5% in its overall production. The Petroleum
Gazette describes the situation in this way ₍₅₎:

"No longer is there any certainty that
Middle East producers will increase
production to meet world demand.

Even Saudi Arabia, which has sharply
increased production at various stages
to counter shortfalls from other OPEC
members' actions, has said importers
can no longer rely on that continuing.

And Saudi Arabia is the world’s
biggest oil reservoirs. "

5. Petroleum Gazette December 1979, page 82.

-195-

The same publication refers to Saudi Arabia's
admonition to the United States to reduce consump-
tion. It says ₍₆₎:

"Saudi Arabia, as its 'moderate’ stance
in OPEC politics has gradually hardened,
has repeatedly called upon the United
States to curb its demand for imported
oil. Unless the United States does so,
then the outlook is for some stern
measures by the Saudis, with possible
repercussions in Australia and other
importing nations too."

There are a number of reasons why the OPEC nations
may restrict supply. First, in a seller's market
the higher price for oil produced is likely to
diminish any shortfall in profit brought about by
restricting supply. Secondly, the OPEC nations
recognise that their oil supplies are finite and
are, inexorably, diminishing. Thirdly, the asset
may be appreciating rather more quickly in the
ground than the profits can appreciate once the goods
have been produced. The Transport and Energy Overview
says this ₍₇₎:

"Since the rapid price increases in
1973 and 1974 OPEC, as a group, have
(sic) found it difficult to adequately
invest their huge surplus revenues...
The difficulty of finding profitable
investment opportunities coupled with
the rapid inflation rate has led to
the situation where a possible strategy
for OPEC, with maximisation of long term
income as a goal, would be to leave the
oil in the ground where it is 1ikely to
appreciate at a faster rate than the
rate of return on the available investments.”

4.6.10 Australia’s Supply of Oil is Highly Uncertain

To the extent that Australia is dependent upon
overseas reserves to meet its transportation needs,
its capacity to fulfil1 those needs must be open to
doubt.

6. Petroleum Gazette December 1979, page 82.
7. ibid., page 39.

-196-

In the United States the Central Intelligence Agency
(C.I.A.) made the following assessment:- ₍₈₎

"(It) believes that world demand for oil
will approach productive capacity by the
early 1980s and will substantially exceed
capacity by 1985. They estimate world
demand for OPEC oil to be between 47 and
51 million barrels a day by 1985 and that
if all other OPEC members produce at
capacity, satisfying this demand would
require Saudi Arabia to produce between
19 and 23 million barrels a day. Given
the present Saudi capacity of 10-11
million barrels a day and the C.I.A.'s
assessment that the Saudis are two years
behind with existing expansion plans the
C.I.A. concludes that world demand will
exceed productive capacity from a practical
viewpoint."

Similar views have been expressed by the Joint
Committee on Foreign Affairs and Defence and by the
National Advisory Committee ₍₉₎. The Transport and Energy
Overview concluded its analysis with the following:- ₍₁₀₎

"The world has entered an era of higher
priced energy and, in particular, higher
priced oil. The world oil supply outlook
is also uncertain.

-        Various projections of oil supply and
demand point to a supply shortfall
between the mid-1980's and the end of the
century.

-        Since OPEC controls 50% of world oil
production it will determine the future
supply and price of oil.

-        This leads to great uncertainty about
the future world supply and price of
petroleum fuel."

The Chairman of the National Energy Advisory Committee,
Mr. G.J. Lynch made the following comment in Apri1,
1979:- ₍₁₁₎

8. Transport and Energy Overview, page 60.
9. Transport and Energy Overview, page 61.
10. ibid., page 189.
11. The National Energy Scene April 1979, Conference of
the Institute of Engineers.

-197-

"If the situation of 60% dependence on
overseas imported crude oil is reached
by 1987/88, we have a potential problem
of security of supp1y...The producing
countries may possibly decide to limit
their production..there may be a complete
embargo or restriction, at least, for
political or military reasons."

The Petroleum Gazette in April 1979 said this ₍₁₂₎:

"What emerges is that the level at which
OPEC Governments are willing to produce
oil will be more important than (other
factors)..

It is an understatement to say that
there are uncertainties in the world
oil supply situation."

In the December edition of the same gazette the
following is said ₍₁₃₎:

"There is a new dimension in the interna-
tional oil market. A year ago the
dominant question was ‘what price?’.
More recent events make the most pertinent
question for Governments and industry:
‘What supplies?'.

But availability of oil, at any price,
is of paramount importance in the new
energy equations. A new term, 'discre-
tionary oil', has come into the reckonings."

The Study Group appreciated this uncertainty only too
we11. It said ₍₁₄₎:

"FIELD: I think the single one (factor)
that is most unknown is the supply of
fuel. Everything else actually is
probably more stable. The key variables
which we consider are in some ways more
stable than they have been in the past."

In another session the following was said (15):

"FIELD: The real argument is not so much
about price: no one is very worried
about price; it’s the availability
question. That’s an issue which is just
completely unknown."

12. Petroleum Gazette April 1979, page 15.
13. Petroleum Gazette December 1979, page 75.
14. Transcript UTSG (now STSG), 12/12/79, page 45.
15. Transcript UTSG (now STSG), 4/12/79, page 96.

-198-

So where does that leave us? There is the prospect
of declining self-sufficiency by 1988 in the absence
of new discoveries. There is the long term prospect
(well into the 1990s) of a substitute fuel. There
is a likelihood, in the meantime, of some dependence,
great or smal1, upon imported oil from the Middle
East. There is, therefore, a susceptibility, to
disruption in supply.

What are the implications for this Inquiry? One
witness suggested that in 20 years’ time the road, if
built, would be useful to children using skateboards ₍₁₆₎.

We feel the conclusion offered by the Total
Environment Centre (and Friends of the Earth) is not
unreasonable. It says ₍₁₇₎:

"Australia is faced with an increasing oil
supply-demand gap. While in the long
term this gap will be narrowed by the use
of synthetic and the development of new
transport technologies, it appears that
the age of cheap transportation fuels is
over, and that Australian society will
have to adapt its social, economic and
spatial relations to an age of costly
and less plentiful fuels."

The Inquiry, in short, must view the traffic demand
projections with a conservative eye.

4.6.11 The Effect of Fuel Price Rises on Travel
Demand

The Bardwell Park Community Resources Centre made
the following submission, having conducted a survey
in the area ₍₁₈₎:

"The replies (to the survey, and at a
group meeting) are very informative.
The question referred to ‘a doubling
of petrol prices tomorrow' suggests
that demand is not cost inelastic
because virtually nobody would use
their cars more, or use public transport

16. Transcript 2/10/79 page 39, Mrs. Phillips.
17. S.K/C 1287 Submission Total Environment Centre/
    Friends of the Earth page 17 and see Submission
    S.K/C 341 Canterbury Municipal Council, page 16.
18. S.K/C 950 Bardwell Park Community Resources
    Centre Attachment A.

-199-

less. Many people also indicated they
would use their cars less and fewer
people thought they would use them
the same amount."

The proof of the pudding is in what people do rather
than what they say they will do. In Europe, where
the price of petrol is considerably higher than it
is in Australia, it appears that the consumer has
switched to sma1ler, energy-efficient cars, than was
the case before the price rise. Travel, as such,
has not appreciably diminished ₍₁₉₎. The same
response was detected in the United States after the
'energy crisis' in 1973 ₍₂₀₎. Further, it appears
that the economies made by the consumer are in off-
peak trips for recreation or holidays or shopping ₍₂₀₎.

The Department of Main Roads, in its submission,
referred to the inelastic nature of travel demand.
It said ₍₂₁₎:

"Evidence available from the 1973 'energy
crisis' which reduced petrol supplies and
resulted in a sudden 50% price increase
in the United States shows that car travel
was not substantially reduced, nor was use
of public transport substantially increased.
Reduced availability of petrol appears to
have accelerated a trend to smaller and
more fuel efficient vehicles, not a switch
to other modes of travel.

Petrol price elasticity, defined as the
percent change in car use caused by one
percent change in petrol price, was found
to be very low..taking -0.1 as the average,
an immediate increase of 50% in the price
of petrol would reduce car usage by only
5% for the average working day."

The Department urges caution in the application of
the United States experience. Sydney, and the travel
habits of its populous, may be different to the
United States. The price of petrol, even at the

19. The Future of the Car in Urban Transport, D.A.
    Hensher, page 12.
20. Impact of the Energy Shortage on Travel Patterns
    and Attitudes, John F. Sacco (Transport Research
    Record-561), page 1.
21. DMR Submission S.K/C 340 De Leuw Cather, September
    1979, page 22.

-200-

present time in the United States, is extraordinarily
low by world standards.

It does appear, however, there is some prospect of
increasing vehicle occupancy as a result of higher
petrol prices ₍₂₂₎. Vehicle occupancy in turn will
effect the number of vehicles. It furnishes a
further reason for adopting a conservative view
of estimates of future traffic demand.

4.6.12 Postscript on Land Use

Conservation is obviously called for. Various
strategies have been suggested. The Transport and
Energy Overview has this suggestion ₍₂₃₎:

"Land use changes have probably the
greatest long term potential for
limiting the demand for transport
fuel. For example, on the basis of
extensive modelling work..the Urban
Institute is confident that attitude
change, in favour of living near work
or in favour of living near neighbours,
would result in very large energy
savings. The study concluded that only
when lifestyle and living pattern
changes were included in the analysis
could substantial decreases in per
capita energy consumption in urban
transport be detected. By 2000 living
close to work showed up as a somewhat
more effective way of saving transport
energy than increasing residential
density."

This passage underlines the need to carefully weigh
the land use implications of facilities which are
proposed. Will the Cooks River Option result in
cross-regional commuting and prejudice a policy of
regional self-containment? Will the South Western
Option promote further urban sprawl?

22. DMR Submission S.K/C 340 ibid., page 22.
23. Transport and Energy Overview, page 151.

-201-

5. THE APPLICATION OF THE MODEL IN THE PRESENT INQUIRY

5.1 The Triptables Used in the Inquiry

In the course of the Inquiry certain errors in the
triptable were detected. The errors were minor,
but they had the effect of distorting the number
of trips between the South-West of Sydney and the
Warringah Peninsula. For convenience the trip
tables were described in the following way:

Trip table A was the triptable
which contained the error. The
triptable was not used in the
Inquiry except (apparently) in
the calculation of certain
pollution and noise levels.

Triptable B was developed by
consultants to the Department of
Main Roads, De Leuw Cather Australia
Pty. Limited, in connection with a
study of the South-Western Freeway
(completed in September 1978). The
consultants (and the Study Group)
realised there was an error. A
manual adjustment was made to
compensate for the error. Trip-
table B played no real part in
the present Inquiry.

Triptable C was developed by the
Study Group for the purposes of
the Joint Study Report (prepared
by the Department of Main Roads
and the Planning and Environment
Commission). The Joint Study Report
was used by this Inquiry as a back-
ground document. It was widely
disseminated. The error in triptable
A was corrected by applying a factor
to certain trips. There remained a
'bug' in the South-Western Region.
For the purposes of the economic

-202-

analysis the jurisdiction (being a
collection of zones) thought to be
effected by the 'bug' was simply
omitted (jurisdiction 14) .

Triptable D was developed when it
was appreciated that triptable C
was not adequate. The factoring
process applied to triptable A
introduced certain distortions to
the projected travel. The model was
re-run (at least in part). This
triptable was used in the final
submission by the Department of
Main Roads (embodied in the Report
of its consultants, De Leuw Cather
Australia Pty. Limited) in September
1979) ₍₂₄₎.

Analysis has been complicated by the need to
differentiate between triptables.

5.2 Can the Inquiry Rely Upon Triptable D?

The Department of Main Roads, the State Transport
Study Group and the consultants to the Department
of Main Roads, De Leuw Cather Australia Pty. Limited,
each expressed the view that triptable D was reliable.
Mr. Haines from De Leuw Cather said this ₍₂₅₎:

"HAINES: Triptable D was developed
shortly after we were briefed (by the
DMR) but we did not initiate changes
to triptable C that produced trip-
table D and we were not familiar with
triptable C. We saw some trip end
summaries of triptable D when it was
produced and nothing in the summaries
alerted us to any problems.”

The Department of Main Roads in an earlier submission
had referred to 'a calibration problem' at the eastern
end of the major options (i.e. in the vicinity of the

24. S.K/C 340 Department of Main Roads Submission.
25. Transcript DMR, 10/12/79, page 78.

-203-

airport). The consultants were asked about this:

"COMMISSIONER: Did you not perceive
that there was a calibration problem
at the eastern end of the network,
that is to say in the region broadly
from General Holmes Drive across to
Undercliffe, arising from triptable D?

HAYNES: We perceived during our work
that the link assignments in that area
did not match closely with counts on
the 1976 triptable but the screenline
crossings did appear to match reasonably
well with counts. If that happens the
problem can either be with the triptable
or it can be with the network. Because
that is a very complicated area, we chose
to assume that the problem was with the
network and that provided we had a
reasonable correlation of trips across
screenlines with counts, that tended to
support our view that it was a network
assignment problem rather than a trip-
table problem."

Mr. Carlisle from the same consultants added this ₍₂₆₎:

"CARLISLE: The same problem occurred
during SATS. Other studies that have
looked at the airport problem have all
reached the same conclusion, that there
is a demand that if the airport wasn't
there would be going through the middle,
and any traffic assignment process has
a large problem in handling a large
lump of traffic that has two alternative
paths that become very close in travel
time and distance. Since 1972 this problem
has continually been one that's plagued
us; it's one that we have learned to live
with, much more than one that immediately
makes us worry about the triptable..."

This is reassuring. The Inquiry, however, has some
lingering doubts. They arise from two things. First,
an exhibit was tendered (Exhibit 66) which compared
the growth rates in the two triptables, triptable C
and triptable D. The exhibit is reproduced below ₍₂₇₎:

26. Transcript ibid., page 29 .
27. Exhibit 66.

TABLE 17

COMPARISON OF TRIPTABLES AND TRIP GROWTHS

Triptable	Total Trips in Triptable
Triptable	(Total Truck)	Truck	Total Vehicle
1976	609,965	34,414	646,379
1991 ‘C’	767,686	40,930	808,616
GROWTH ‘C’			25%
1991 ‘D’	743,555	40,930	784,485
GROWTH ‘D’			21%

It will be seen from this table that the growth rates
are:

25% growth triptable C
2l% growth triptable D

Yet the individual loading on every link (so far as
we can see) for triptable C is less than for trip-
table D. One would have expected the reverse. Since
the growth in triptable D, region-wide, is less than
triptable C one would have expected that the traffic
on each individual link would also be less.

The Inquiry was especially concerned with an area
(known as the Study Area) which extended as far as
King Georges Road. Yet a number of maps with the
individual loadings from triptable C and triptable D
were furnished to the Inquiry ₍₂₈₎. They extend from
the main western railway line to the Georges River.
In that spread of country the link loadings on trip-
table D are consistently higher than triptable C.

The paradox was put to the Department of Main Road.
It was agreed that it ‘just seems odd’ (29). It
referred to the Study Group. It carried out a check
across various screen lines. The result confirmed
the Inquiry's impression that the growth under trip-
table D was consistently higher than triptable C.
Reproduced below are the various tables ₍₃₀₎:

28. Exhibit 102 to 113.

29. DMR Transcript 18/3/80, page 73.

30. Letter from UTSG 22/5/80, page 2.

-205-

TABLE 18.

PROJECTED PEAK PERIOD MOVEMENT (MAJOR DIRECTION)

BASE CASE

SCREENLINE 1 4-5 8-9	COMMITTED ROADWORKS OPTION (ALTERNATIVE G)
	‘C’	‘D’
	15,814	16,569 (+5%)
	18,509	19,403 (+5%)
	19,646	21,209 (+8%)

TABLE 19.

PROJECTED PEAK PERIOD MOVEMENT (MAJOR DIRECTION)

COOKS RIVER ROUTE

SCREENLINE 1 4-5 8-9	COOKS RIVER ROUTE OPTION (ALTERNATIVE A2)
	‘C’	‘D’
	16,063	16,831 (+5%)
	19,522	20,507 (+5%)
	20,723	22,412 (+8%)

-206-

TABLE 20.

PROJECTED PEAK PERIOD MOVEMENT (MAJOR DIRECTION)

SOUTH-WESTERN FREEWAY

SCREENLINE 1 4-5 8-9	SOUTH-WESTERN FREEWAY OPTION (ALTERNATIVE B3)
	‘C’	‘D’
	15,871	16,700 (+5%)
	19,130	20,424 (+7%)
	20,530	22,332 (+9%)

The Study Group makes the following comment upon
these tables ₍₃₁₎:

"It is evident from the above that
across the three screenlines, trip-
table 'D' has between 5 and 9 percent
more trips than triptable 'C' which
does not represent a significant
difference, and is within the
anticipated accuracy of the modelling
process. Nevertheless it is important
to note that for the purposes of
comparative evaluation of the major
road options, that the relativity
between the options is the critical
factor."

The Study Group then points out that the relativities
are ‘absolutely consistent irrespective of which
triptable is employed'.

We concede that the difference is not great. We
acknowledge that the relativities do remain constant.
Yet we are nagged by the paradox of individual links
showing lower traffic loadings under triptable ‘C’
and yet, region-wide, showing a greater growth in
traffic.

31. ibid., page 2. An explanation not advanced in the
    letter but subsequently suggested by a Study Group
    member is that the 'bug’ in C was responsible for
    the additional trips.

-207-

The second source of our puzzlement concerns the
assigned loadings on individual links. In some
cases, as a matter of commonsense, we would have
expected particular roads to show an increase in
traffic. Yet they do not. Figure 6.1 in the
Joint Study Report depicts the significant changes
in traffic levels brought about by the Cooks River
Option when compared to the base case. Between
Wardell Road and General Holmes Drive there are no
significant increases in traffic apart from the
section of Wardell Road itself between the Cooks
River Route and Ewart Street, Marrickville. We
should have thought that there would be a significant
increase in traffic in the following:

Illawarra Road
Carrington Road (from which there is
a spur giving direct access to and
from Carrington Road/Victoria Road)
Unwins Bridge Road
Princes Highway itself (which is
shown as a significant decrease)
alternatively Marsh Street (which
is not affected)
alternatively General Holmes Drive
(which shows no affect).

Taking the cluster of roads Princes Highway, Marsh
Street and General Holmes Drive, there are a number
of factors. First, there is the calibration problem
to which Mr. Carlisle referred. Secondly, the
significant difference is defined as a 10% change.
To make a significant difference to General Holmes
Drive there must be significant additional traffic.
Yet, even making allowance for this, it still seems
odd to us that neither Illawarra Road nor Carrington
Road nor Unwins Bridge Road nor Marsh Street nor
General Holmes Drive do show a significant change, and
Princes Highway shows a significant decrease. The
commonsense expectation we had was apparently shared
by those who prepared the Joint Study Report. In the
summary to that Report the following appears referring
to the Cooks River Option ₍₃₂₎:

32. Joint Study Report DMR/PEC, Page 1.

-208-

"Congestion: some at ends only"

In the body of the Report the following is said (33):

"Some congestion would be experienced
on the roads giving access to each end
of the route, however, such as General
Holmes Drive and the Hume Highway."

Certainly it may be suggested that these passages are
speaking of something different. The diagram is
referring to a significant change in traffic levels,
whereas the Report is referring to ‘congestion’.

The Joint Study Report does refer to the level of
traffic, however, in South-Eastern Marrickville.
It says ₍₃₄₎:

"Further, it is to be expected that
the connection from the Cooks River
Route at Bayview Avenue to Carrington
Road would increase traffic loadings
on roads in the south-eastern area of
Marrickville, such as Unwins Bridge
Road, Carrington Road and Victoria
Road."

Yet Carrington Road and Unwins Bridge Road are not
shown as experiencing significant changes in traffic
levels compared to the Base case.

In this respect it is interesting to compare the
comments of the Study Group made in the course of
the Central Industrial Area Study ₍₃₅₎:

"Construction of the Chullora/Kyeemagh
route will increase the traffic demands
on Unwins Bridge Road. The limited
capacity of Unwins Bridge Road at Tempe
will divert some of this excess demand
onto parallel back streets, such as
Carrington Road, while Unwins Bridge
Road will be saturated for most of the
morning peak period."

33. ibid., page 26.
34. Joint Study Report, page 26.
35. Central Industrial Area Study UTSG (now STSG) page 65.

-209-

The Study Group was there testing a freeway version
of the Cooks River Route, and was using a different
triptable (and indeed a different network, especially
developed for the purposes of the study). Nonetheless,
their comment embodies something which one would
expect intuitively. If there is a direct spur
leading into Carrington Road the traffic on that
road must surely increase. At the very least it
must increase in the roads which are parallel on
either side (Illawarra Road and Unwins Bridge Road).
None of these roads, as we say, demonstrate such an
increase.

There are other illustrations. We do not doubt that
the model is of enormous assistance in appreciating
where reductions and increases in traffic are likely
to occur. We simply reaffirm that the suggested
traffic flows are a very general guide to what may
happen. What they suggest may be pushed and pulled
in this direction and that, aided by (for want of a
better term) a certain degree of commonsense. The
Study Group, we might add, would not suggest otherwise.

5.3 The Model Measures Corridor Movements

The Study Group repeatedly stressed that the loadings
on individual links were indicative only. The model
was designed to reproduce a corridor of movement
rather than individual flows on particular streets.

Yet parties making submissions to the Inquiry, on
occasions, tended to use the model beyond its capacity.
Very specific increases and decreases were suggested
for individual links. Professor Beesley puts the
point in this way ₍₃₆₎:

"The critical methodological point,
perhaps not appreciated fully in the
hearings, is that in order to answer
the question about investment in any
given possible road network, the level
of detail of description of flows has

36. Annexure 2 'Cost Benefit Analysis and the
Kyeemagh/Chullora Road Inquiry’, M.E. Beesley
page 362 of this Report.

-210-

to be greater than that for which the
answers are sought. The networks used
as descriptors in the Inquiry were
developed in order to illuminate invest-
ment at a higher level - e.g. priorities
in major corridors. But somehow the
need to have any information about the
cases at hand overcame methodological
good sense. One then had the absurd
spectacle of debates about expected
flows on particular links of the network,
how they were individually expected to
grow, etc. At this level, considerable
uncertainty is quite acceptable in
principle so long as aggregation over
observations is acceptable. For the
original model purposes, it was; for
the Inquiry purposes, it was not."

Professor Beesley then suggests an alternative ₍₃₇₎:

"If one is to retain the basic model,
one has specifically to 'window’ an
area of the size covered by the Inquiry,
and develop accounts of flows likely to
come across the 'window frame’, concentr-
ating attention on an additional and
detailed model of movement within.
Alternatively, one could quite reason-
ably have taken the view that traffic
as a whole in the most relevant area,
was not really likely to change much
anyway, except in composition, over the
foreseeable future. One could then have
concentrated on considering the effects
of a range of possible moves to combat
environmental deterioration and, just
as important, to improve the experience
of the population at present exposed."

5.4 The Use of the Model in the Inquiry

The exposition based upon the model concentrated on
two things:

the traffic passing through certain
shopping centres
the traffic filtering through
residential areas.

Both were perceived, rightly, as important community
problems.

37. Annexure 2 ‘Cost Benefit Analysis and the Kyeemagh/
Chullora Road Inquiry', M.E. Beesley, page 363 of
this Report. (paragraph 24)

-211-

Yet the use of the model, and the traffic flows it
represents, to draw inferences concerning either of
these matters is suspect. The traffic model
reproduces the traffic flows for the morning peak
period between 7.15 a.m. and 9.14 a.m. Needless
to say, people do not shop during those hours.
Whilst traffic engineers have a rule of thumb, that
the morning peak traffic represents a certain
proportion of the daily traffic on a particular road,
it is but an approximate guide. One would have to
allow for a wide margin of error. The composition
of traffic in a particular location may change
markedly during the course of the day. We venture
to suggest that if containers are not railed to the
Western Suburbs certain roads will take on a
different character during the off-peak than during
the peak, whatever the ratio between peak and off-
peak traffic may be. Yet the Joint Study Report,
and various other documents, used the model to
suggest highly specific reductions (or additions)
to the traffic flow through shopping centres. These
were hailed as a triumph (or defeat) for one option
over another. We doubt that the model can be so
employed.

In the case of residential streets, it will be
remembered from our description of the model network
that local streets are omitted. The 11,000 links
comprising the network only include major roads and
a few strategic local roads where they carry a
significant traffic volume. Certainly these roads
in many cases (and for much of their length) are
lined with ordinary homes and are, therefore,
‘residential'. Nonetheless they form by and large
the arterial road network. That network is hardly
a reliable basis, in the Inquiry's judgement, upon
which to draw an inference concerning the extent to
which traffic is attracted away from 'residential
areas’.

-212-

VIII PRIORITIES

1. THE ROAD CONSTRUCTION BUDGET

1.1 The Road Construction Budget is Shrinking

There is, each year, in the Annual Report of the
Department of Main Roads, a lament that its budget
is less now than it was ten years ago (when the
traffic was less than it is today). The most recent
Annual Report says this:- ₍₃₈₎

"The real value of revenue made available
to the Department has fallen considerably
since 1972/73...When expressed in 1978/79
values (funds) have declined by 18%:
from $339,000,000 in 1972/73 to $279,000,000
in 1978/79."

The reduction in funding for the County of
Cumberland is depicted in the following graph:- ₍₃₉₎

FIGURE 14.

EXPENDITURE ON CONSTRUCTION OF ROADS AND BRIDGES IN
THE COUNTY OF CUMBERLAND BY D.M.R. 1949/50 TO 1976/77

38. D.M.R. Annual Report 1978-79, page 8.
39. The Sydney Transport Planning Myth, K.W.
Dobinson, May 1978, page 6.

-213-

The gulf between present road funding, and that
envisaged by the Sydney Area Transportation Study,
emerges from this diagram.

There has been a re-ordering of priorities in terms of
Government spending. The budget for Social Welfare,
Health and Education, has increased in real terms. Road
building has decreased. The extent to which there has
been a 'squeeze’ upon roads emerges from the following
table:- ₍₄₀₎

TABLE 21.

COMMONWEALTH GOVERNMENT EXPENDITURE
1969/70 AND 1978/79

Function	1969-70	1978-79 (a)	Average Annual Growth Rate %/p.a.
Education	244	2498	29.5
Health	462	2913	22.7
Social Security and Welfare	1269	8015	22.7
Roads	218	553	10.9
Payments to States, N.T. and local government	1757	6557	15.8
All Other	3398	8334	10.5
Total Budget Outlay (c)	7348	22870	16.4
Total Commonwealth Outlays (c)	7542	30088	16.6

(a)    Estimates (see source for this Table).

(b)    Tax sharing, net other general revenue assistance,
net State Government loan programs, assistance related
to State debts, local government tax sharing
entitlements etc., net natural disaster relief,
payments to the Northern Territory.

(c)    For definition see the Appendix to the Budget
Statements (The Functional Classification of Budget
Outlays) and pp. 194-195 of Statement No. 6, Budget
Speech 1978-79.

Source: Treasurer of the Commonwealth of Australia,
Budget Speech 1978-79, AGPS, Canberra, 1978.

40. An Assessment of the Australian Road System 1979
(Part 1), Bureau of Economics, page 31.

-214-

1.2 The Present Sydney Metropolitan Budget

In the year 1978/79 the Department of Main Roads had
at its disposal the sum of $355,000,000 (in round
figures) for road building. That sum was divided into
the following parcels:- ₍₄₁₎

$84,000,000 to the County of Cumberland
$145,000,000 for the country
$67,000,000 for National Roads (a
Commonwealth Road Grant)
$20,000,000 for traffic facilities
$44,000,000 for various other projects

(the amount has been rounded)

From the $84,000,000 available for the Sydney
Metropolitan Area approximately $23.5 million must be
devoted to maintenance of the existing road system ₍₄₂₎.
Of the remaining $60 million, a proportion is a
Commonwealth Grant. It is earmarked for the particular
purpose for which the Grant was given. It is not
within the power of the Department to divert that money
to some other purpose which it considers more worthy
(although the Department is invited to make submissions
to the Commonwealth before grants are given, identifying
matters of priority). ₍₄₃₎

The Department, in short, has very limited construction
money at its disposal. It must use its funds sparingly.
It is only too conscious of this limitation. In its
submission the Department says:- ₍₄₄₎

"There has been for a number of years,
a continuing decline in funds available
to the Department for expenditure on
urban road construction, so the annual
rate of expenditure on arterial roads
in the Sydney Area has been far below
that required to overcome the large
backlog of improvements which are

41. D.M.R. Submission S.K/C 340 Document entitled
    'Works Programmer Formulation in the D.M.R.',
     December, 1978.
42. Transcript D.M.R. 10.1.80, page 7.
43. See An Assessment of the Australian Road System 1979
    (Part 2) Bureau of Transport Economics, pages 5-33.
44. D.M.R. Submission S.K/C 340 July, 1979, page 6.

-215-

urgently needed. It seems unlikely
that this position will change in the
near future."

The Planning and Environment Commission say much the
same thing:- ₍₄₅₎

"..Federal policy of funding for major
urban growth and change, including
for urban arterial roads, has been
reversed since the mid-1970s and is
now almost entirely the State’s
responsibility. This is unlikely to
change in the medium term (i.e.,
within ten years) and therefore funds
will be very limited allowing only
the most urgent problems to be addressed."
(emphasis added).

2. THE ORDERING OF PRIORITIES

2.1 The Issues to be Addressed

We have been at pains to stress that a traffic problem
need not be answered by a transport solution. In
every case one must ask the questions:-

What are the land use implications of
doing something?
What are the implications of doing
nothing?
Is it, therefore, desirable that, something
should be done?
If it is, should that ‘something' be a
transport solution (such as the creation
of more road capacity) or a public transport
solution, or a land use solution?

Assuming these questions have been addressed, and
that a transport solution is thought appropriate, there
is a further series of questions which must be
answered: -

45. Planning and Environment Commission submission
S.K/C 947, Appendix 7, page 1.

-216-

Is the need so pressing (compared to other
perceived needs throughout the Metropolitan
Area) that scarce money should be devoted
to finding a 'solution'?
Assuming there is a need, and, assuming it
is pressing, is the preferred solution
disproportionate in cost to the problem
which it addresses?
If the problem is essentially local, and the
cost of solving it is high, does that throw out
of kilter the even spread of scarce funds
throughout the entire Metropolitan Area?
Are there other ways which are less ccstly
(and perhaps more equitable) of solving the
problem?

2.2 The Method Employed by the Department of Main Roads

The Department, quite rightly, sees its first
responsibility as the maintenance of the existing
road system. It draws a distinction between maintaining,
sustaining and extending the road network ₍₄₆₎:-

“CROWE: We, in fact, have an aging road
system.. Our first task is to maintain
it.. Which is basically just keeping
the system running. What I call sustaining,
is that we have a system which we
have to renew and rehabilitate...
We actually look at improving it and
we improve it for two reasons. We improve
it because there is increased demand on
the system. We, in fact, improve it also
to improve the service that we are offering
to the public."

Mr. Crowe then gave the following analogy:-

"CROWE: If you look at a taxi fleet,
what do you do to it? In fact, you
maintain it. You grease and oil

46. Department of Main Roads Transcript 10.1.80, page 8.

-217-

change, right? What they do also is
sustain that fleet. They, in fact,
replace the units after 40,000
kilometres. To meet increased demand,
what they do is add to the stock of
vehicles in the fleet. If in fact they
want to improve the service, they provide
better vehicles...There aren't too many
vintage or veteran taxi cabs around.
While we don't have too much vintage road..
we have a lot of veteran roads, and this
is of continuing concern to us."

The Department quite deliberately has pursued, a policy
of decentralising its operations. Its engineers are
located in the field. They are required, by the Main
Roads Act, and as a matter of commonsense, to liaise
with Local Councils. They do so. They are familiar with
the problems within the area which they service. In
the Sydney Metropolitan Area there are two geographical
divisions (the Metropolitan Division and the Parramatta
Division). There are two further Divisions concerned
with major works (the Inner Freeway Division and the
Outer Freeway Division).

The way in which priorities are ordered is described
in the following paragraph:- ₍₄₇₎

"CROWE: We try to provide the Divisional
Engineer with a realistic estimate of the
funds that will be available and say to
him that, "Look, here is our best estimate
of the funds we will have available for
you for works in this particular category
in the next financial year. Please submit
your proposals by the end of January. Your
proposals should not exceed the funds
we have indicated plus 10%". This then
provides a bit of flexibility. If there
is a particular addition to funds we
know where the Divisional Engineer would
like it, spent."

In the nature of things there are certain works in
progress. They must be included in any programme. ₍₄₈₎

47. D.M.R. Transcript 10.1.80, page 19.
48. D.M.R. Transcript 10.1.80, page 20.

-218-

There are other projects which have been under
contemplation for some time. There are others which
represent commitments made by the Government of the
day or the co-ordinating authority, TRANSAC:- ₍₄₉₎

"CROWE: The Department is a responsible
arm of Government and its activities
must follow the policies of the Government_
of the day and I certainly have no quarrel
with that. In the development of its
projects and the proposals (incorporated
into its programmes), the things it does
must be in accord with the Government
philosophy..and I would include the URTAC
(now TRANSAC) philosophy in the broad
context of political commitment".

2.3 The Use by the Department of Main Roads of Cost Benefit
Analysis

In a later chapter of this Report we discuss the
technique of cost/benefit analysis ₍₅₀₎. We were
interested to know the extent to which this technique is
used by the Department in ranking various projects. The
answer emerges from the following exchange:- ₍₅₁₎

“CROWE: Well the short answer I suppose,
is that the Department has not used benefit-
cost ratios to a great extent in the sense
that, shall we say, a ratio is automatically
calculated for each project. We have used
(them) selectively in the development of
some projects. I am thinking particularly
of a study we did of Parramatta where, in
fact, we did develop benefit cost ratios
for various alternative proposals that we
had in the Parramatta area.

COMMISSIONER: So in other words you identify
the problem area and then you...endeavour
to rank solutions..

CROWE: We did in fact rank solutions
on me basis of cost/benefit analysis.
That’s right...I think it would be fair
to say we are using it more.”

49. Transcript ibid., page 45.
50. See page 259.
51. Transcript ibid., page 39.

-219-

2.4 The Equitable Allocation of Funds

The road system is rather better in some areas than
it is in others. There is an expectation on the part
of the community, nevertheless, (and especially their
elected representatives in Local Councils) that a
reasonable share in the road funding available will
be allocated to each area. If they perceive
neglect they feel it keenly. The following
submission was made by Mr. Vincent Durick, M.P., the
Member for Lakemba in the Legislative Assembly of
New South Wales:- ₍₅₂₎

"I feel that there are (certain) matters
which should be taken into consideration:

1. The comparative neglect by the
   Department of Main Roads of the
   Canterbury-Bankstown area during
   this last 15 years."

Mr. Durick supported this statement with the following
evidence:- ₍₅₃₎

"It is only in the recent period that
any great activity has been shown by
the Department of Main Roads in the
Canterbury-Bankstown area... On the
16th October, 1968 the Minister for
Local Government and Highways in answer
to Mr. R.J. Kelly, who at that time
was the member for East Hills in the
State Parliament, gave information in
regard to a question seeking information
regarding the expenditure on main roads
funds North and South respectively of
the Harbour and Parramatta River during
the previous 3 years. In answer the
Minister indicated that from 1st July, 1965
to the 30th June, 1968 $M37.938 had been
spent North of the Harbour while at the same
time only $M10.858 had been spent South
of the Harbour.

The amount expended North of the Harbour
included $M22.933 for the Warringah Expressway
and the Newcastle Tollway South of the
Hawkesbury River.

52. S.K/C 201, Mr. Vincent Durick letter 29.6.79
53. Ibid., pages 1 and 2.

-220-

But even allowing for the subtraction
of this amount, the amount expended North
of the Harbour is almost 50% greater
than that expended in the Southern area.

As a conservative estimate there would
be nearly three times as many people
living in the relevant Southern area.

Unfortunately, the Canterbury-Bankstown
area received very little of the amount
spent in the Southern section."

We shall refrain from commenting upon the substance
of this submission beyond saying that the figures
now more than a decade old. The submission does
underline, however, (as do many others) that the
community is conscious of the need for scarce road funds
to be scattered evenly throughout the entire Metropolitan
Area.

The Department of Main Roads is also conscious of that
need. Indeed, it threw its considerable weight behind
the South Western Option because it would spread benefits
far more widely across the Southern metropolitan area
than the Cooks River Option. It said:- ₍₅₄₎

"..It is seen as the responsibility of
the Department to take into account
factors outside the geographic area
of the Inquiry's specific interest and
to examine the route options as they
interact with overall, metropolitan-
wide needs.”

Later in the same submission it says:- ₍₅₅₎

"Further, no assessment was made of the
effects that a concentrated expenditure
of construction funds on any Kyeemagh/
Chullora Route may have by reducing the
funds available for planned construction
in other areas of Sydney.”

54. D.M.R. Submission S.K/C 340, July 1979, page 3.
55. Ibid., page 5.

-221-

Emphasis is given to the same point by the following:- ₍₅₆₎

"Bearing in mind the constraint imposed
by limited funds available for major
road proposals, it is important that
due regard be paid to the overall
regional benefits which are likely to
be derived rather than undertaking
heavy expenditure for relatively
localised community benefits.”

The Department characterised the Cooks River Option
as a ‘lower traffic demand corridor' ₍₅₇₎. It
concluded its commentary upon this option with these
words (specifically directed to the equity issue) ₍₅₈₎:-

"The high cost of the work (i.e., the
Cooks River Option) therefore, when
considered in relation to the relatively
small part of the Sydney Metropolitan Area
which would benefit, is of particular
concern to this Department.”

This is a matter of some importance. Ultimately, in
the evaluation of options, we will have to examine how
bad the congestion is within an area which may be
relieved by building one or other of the options
(assuming it is thought as a matter of land use policy
the congestion is best eliminated). We will examine
how poor accessibility is from the viewpoint of commuters
and truck drivers. We will express a view as to whether
the expenditure contemplated (in excess of $50 million)
is disproportionate to the problems revealed by the
evidence.

2.5 The Need to Examine a Range of Alternatives

Where a transport solution is thought appropriate to
a traffic problem, it may take one or a number of
forms:- ₍₅₉₎

56. Ibid., page 6.
57. Submission ibid., page 12.
58. Submission ibid., page 12.
59. Alternatives for Improving Urban Transportation,
October, 1977, page 2-1.

-222-

Providing additional facilities
Expanding existing facilities.This can be done by:-

o   road widening

o   widening intersections (since
intersections are usually the
weak link in the chain)

o   adding right-hand turning bays
to assist the capacity of the
intersection and the flow of
traffic

Improving the efficiency of existing
facilities. This may include:-

o   traffic management measures
clearways, priority roads,
way systems etc)

o   preference to high occupancy vehicles
(e.g., bus only lanes, transit lanes;

o   the installation of park and ride
facilities at railway stations and
other inter-changes.

The Inquiry should be in a position to consider each
alternative. The following was suggested by NAASRA:- ₍₆₀₎

"A range of alternative solutions is
thus possible to most transport problems.
The following generally need to be
considered: -

Alternative scales of improvement

Alternative rates of improvement

Improvements to alternative modes

Regulatory and licensing alternatives

A land use change

Combinations of the above

Making no improvement.”

Almost the same list appears in another publication,
interspersed with the following comments:- ₍₆₁₎

"Planning bodies need the capacity to
consider a wide range of alternatives

60. Review of Transportation Planning in Australia
    1977, page 15 and Appendix E, page 2.
61. A Discussion Paper of Transport Planning in
    Sydney to 1975, Rattray and Sinclair, Bureau of
    Transport Economics (Occasional Paper No. 16), page 86.

-223-

to the solution of particular problems.
In the transport field, a wide range
of alternatives may be applicable, as
transport is seldom an end in itself...
seldom is such a range of possibilities
canvassed and in many cases inefficient
use of public funds has resulted."

What happens in other road inquiries? The practice in
Melbourne emerges from the following comment by its
Chief Planning Engineer, Mr. R.T. Underwood:- ₍₆₂₎

"In recent times a number of corridor
studies have been carried out in metropolitan
Melbourne. These studies have examined
existing and estimated future travel
requirements in the particular corridors,
and have assessed the consequences of a
range of possible alternative courses of
future action such as traffic management
schemes, spot improvements and major road
(freeway and/or arterial schemes), together
with the do nothing alternative. As
appropriate, they have also considered public
transport proposals."

A number of illustrations are given. It appears that a
range of alternatives were considered in the following
Melbourne studies:-

The Eastern Corridor Study
The Gardiners Creek Valley Study
The Outer Ring Study (Diamond Creek
to Ringwood)
The Bell Street - Banksia Street Study.

Small improvements can be effective. In its examination
of the Cooks River Option (the Kyeemagh/Chullora Route
designed to Freeway standard) the Study Group made the
following comment:- ₍₆₃₎

"If anything, the analysis of critical
intersections showed the need for a

62. New Urban Roads - A case for Keeping Future Options
Open, R.T. Underwood, page 142.
63. UTSG (now STSG) Central Industrial Area Study,
June 1977, Chapter IV, page 26.

-224-

large number of small, localised
improvements within and immediately
surrounding the industrial area, rather
than a small number of large new roads.”

We are not suggesting that the Department of Main
Roads is not aware of the need to examine a range
of alternatives. It is. The following exchange gives
but one illustration:- ₍₆₄₎

"DOBINSON: I was very concerned that we
were developing a proposal - or had a
proposal thrust upon us - to upgrade
Victoria Road at some quite huge cost
to give what I saw to be a relatively
small benefit and I was curious to see
if we could achieve almost that benefit
by working on Ring Road 3 which is a
much softer option and get a lot of
other benefits for the same money. This
in fact proved to be true.”

Later Mr. Dobinson said:- ₍₆₅₎

"DOBINSON: But to do the same in the
area at which you're looking is again
a very big task: to look at various
options and see what they do and eventually
find the softest one to achieve whatever
the objective is - whether it's to remove
trucks from residential streets or to
shift the traffic in a reasonable sort of
way.

COMMISSIONER: Yes. But would you agree
with me that it is important that it be
done?

DOBINSON: yes it is important that
it be done. As much for the sensitivity
of the area as the expenditure of funds
because it’s quite expensive whatever you
do and we have limited funds.”

The Inquiry is concerned that a range of options has
not been proposed. The only solutions suggested are
in the nature of major surgery. A band-aid may be

64. D.M.R. Transcript 12.5.80, page 48.
65. D.M.R. Transcript 12.5.80, page 48.

-225-

inappropriate. But something in-between may do the
job.

There were a number of possibilities. They were not
examined in depth alongside the road alternatives.
They ought to have been. One alternative was the
widening of existing roads. NAASRA made the observation
that the public tends to accept projects 'up to a
certain size' objecting only (or most vociferously)
to the very large projects (66). There is a lesson in
that. Smaller projects should be considered including:-

road widening
intersection widening
the adding of right-hand turning
bay at intersections

The following was put to the Department:- ₍₆₇₎

"COMMISSIONER: Do you see merit yourself
in proposals that have been put forward
here as ways in which capacity could be
increased by , for instance, adding right-
hand turn bays to Canterbury Road?

DOBINSON: Yes. There is real merit in
that sort of move."

Since it is recognised that the intersection is the
critical element in the street system which limits
road capacity ₍₆₈₎, the question arises whether the
Department (and the community) would be better off with
a series of smaller projects rather than one large
project. The question was put to the Department:- ₍₆₉₎

"COMMISSIONER: In terms of priorities
does the Department feel that it would
be better off spending money on all those
intersections which are demonstrated to

66. NAASRA 'Community and Environmental Aspects of
Urban Highway Proposals’, page 4.
67. D.M.R. Transcript 12.5.80., page 4l.
68. Alternatives for Improving Urban Transportation
ibid., page 12-1.
69. Transcript D.M.R. 25.10.79., page 48.

-226-

have a “Y” value nearing saturation
than in a larger project...I am concerned
about the relative merits of improving
intersections, if intersections are the
weak link in the chain, on the one hand,
and the building of major roads on the
other.

ANDERSON: It worries us too."

Road widening, in many cases, will be less satisfactory
than building a new road. First, widening a major
road, such as Canterbury Road, is difficult because
of the volume of traffic it carries, even in the off-
peak. It can be done, and has been done successfully in
the King Georges Road.. It is disruptive and expensive.
Secondly, utilities (telephone cables, electricity lines
etc) are placed in the pavement. If the road is widened
they have to be moved. The expense is considerable.
It does not arise (usually) in an undeveloped corridor.
Thirdly, it is enormously expensive to resume property
in a heavily built-up area, especially if there are
shopping centres or industrial establishments lining the
route (as often is the case with an arterial road).

Yet there are also distinct advantages. The road will
already operate as some sort of community barrir. It
may even form the boundary of the municipality. The
severance is not likely to be substantially greater if
an important artery is widened from four lanes to six
lanes. The severance would not arise at all, we venture
to suggest, if the widening is confined to intersections.
Yet the capacity of the road may improve considerably.
The gain in capacity may not be comparable to the
capacity which can be obtained with an entirely new
route:- ₍₇₀₎

"COMMISSIONER: If one were simply to
confine one’s ambitions to a few
broadenings at intersections, adding
storage, then the cost would be considerably
less?

70. Transcript D.M.R. 14.1.80., page 73.

-227-

BUNTON: It depends on the frequency
of intersections which are considered
significant enough traffic constraints
to warrant it. You’d have to do an
exercise...the point to be made is
that even if you spend $50 million on
widening all the critical intersections
along Canterbury Road, say from Bankstown
to Newtown or wherever, your increasing
the capacity of that route by fairly
low levels. In other words, the cost of
effectiveness mightn't be in it in
comparison with cost effectiveness of
providing a completely new road. You
might be increasing the capacity 40 or
50 per cent whereas with a completely
new road, for the same amount of money,
your getting (double the capacity) and
more value for the dollar."

Mr. Bunton went on:- ₍₇₁₎

"We probably generally prefer the low
cost small improvement because it's
something you can implement fairly quickly
without distorting the budget and spread
the available funds through the region."

With hindsight, it is unfortunate that this Inquiry does
not have before it, an assessment of the extent to which
the traffic problems of the study area can be solved by
means other than the building of a major road.

71. Department of Main Roads Transcript 14.1.80,
page 73.

Go to B. Planning Criteria -->

C. Economic Criteria -->

The Kirby Reports

Monday 16 March 2015

Volume II - Report (A. Transport Criteria - VI-VIII)