Content uploaded by John F Odgers
Author content
All content in this area was uploaded by John F Odgers on Jul 14, 2014
Content may be subject to copyright.
1
Have all the travel time savings on Melbourne’ road network
been achieved?
A GAMUT Discussion Paper
John Odgers
School of Management, RMIT University
Abstract
The foremost economic benefit postulated and claimed for all road network investments
is the value of travel time saved. This paper’s aim is to empirically test whether the very
substantial economic resources that have been consumed over the last two or so decades
in the construction and use of major road network additions in Melbourne have helped to
achieve the travel time savings which formed the main foundation of their economic
justification. The study uses the annual traffic system monitoring data prepared by
VicRoads for the monitored urban road network, and compares these actual data against
the results for Melbourne’s urban road system that were projected by various traffic
modelling experts in the 1990s. In particular this study uses City Link as the case study to
enable the comparison between such projections and actual traffic volumes and traffic
behaviour.
One of the key findings of this study is that the average whole of day speed on
Melbourne’s freeways overall has stayed at around the same level (78+ kms /hour) apart
from 2000-01 (83.5 kms/ hour) and 2001-02 (79.5 kms’ hour). Second, the average speed
in kilometres per hour in both the morning and evening peak periods for the whole
monitored urban network in the most recent year for which data are available — the year
ended June 30 2007 — is the lowest it has been since 1994-95. Third, average travel
speeds in inner Melbourne post the opening of City Link have reduced in both the
morning and evening peaks. Even more concerning is the fact that average speeds across
the whole day for both freeways and all types of arterial roads in the inner Melbourne
region have all similarly dropped over the years 2001-02 to 2006-07. Fourth, the
projected volume of freeway traffic of 331,000 DVH by 2011 looks unlikely based on the
reduced speed of freeway traffic volumes since 2003-04 and a total freeway volume of
244,700 DVH in 2006-07. Finally, the Net Present Value (NPV) of the extra travel time
in actual daily vehicle hour equivalents over the years 1997-98 to 2006-07 as compared
to the projected total saving in DVH of 161.2 million is -$349.4 million. Given this
dissaving, the NPV of travel time savings on Melbourne’s’ urban road network from
2007-08 to 2030-31 inclusive would have to amount to $1.834.4 billion, as opposed to
the projected NPV of $957.1 million by Allen consulting group and Cox (1996).
In sum, the results from this study suggest that the core of travel times savings benefits,
which is an increase in average travel speeds, has not to date eventuated in Melbourne’s
urban road network during the years under review. Indeed, based on the evidence
presented and analysed in this paper, one could be led to the conclusion that investments
in Melbourne’s urban road network have resulted in more time being used by
Melbourne’s motorists rather than less time. Hence major road infrastructure initiatives
and the consequent economic investments have not yet delivered a net economic benefit
to either Melbourne’s motorists or the Victorian community. Equally concerning is the
plausible conclusion from this analysis that over their remaining economic life such
major urban road network investments are unlikely to result in major travel time savings.
Introduction
‘We might ask whether researchers have in fact been able to measure the travel
time savings expected to be associated with infrastructure investment, such as a
new or widened road that has been built with the intention of generating such
savings. When I searched the literature – specialist books, journals and data bases
– I was very surprised to draw a blank’ (Metz, 2008, p. 30)
David Metz was chief scientist to the UK Department of Transport and is now visiting
professor of transport studies at University College London. The total kilometres of road
that comprise the urban road network of Melbourne has increased substantially over the
last fifteen or so years. This paper takes up Metz’s challenge. Its aim is to shed some
empirical light on whether the enormous economic resources that have been consumed in
the construction and use of these major road network additions have achieved the travel
time savings which formed the main foundation of their economic justification.
The substance of this paper comprises the following sections:
● Overview of Melbourne’s urban road network;
● Projections and economic justifications for major network projects;
● Comparison and analysis of projected performance against actual
performance;
● Conclusions.
Overview of Melbourne’s urban road network
Melbourne’s urban road network is extensive and constantly being modified and
expanded. Its major elements are managed for the Victorian government by VicRoads. Its
extent as at June 2007 is shown in Figure 1.
2
3
Figure 1 Melbourne’s Urban Monitored Road Network (Source: VicRoads, 2008, Traffic System
Performance Monitoring 2006-07, p. 11)
VicRoads classifies the metropolitan road network in two main ways: first by
distinguishing between freeways and arterials, and also by type of arterial (divided,
undivided, with or without tram lines), second by distinguishing between inner and outer
areas.
Melbourne is served by a number of freeways. The first section of Melbourne’s first
freeway system — the South Eastern Freeway — was opened in 1962 (Lay and Daley,
2002, page 261). In 1965 the Victorian State Government announced a ten-year program
of freeway building (Melbourne Transportation Committee, undated). The 1969
Melbourne Transport Plan further developed this plan; it recommended some 510km of
freeway for metropolitan Melbourne as its ultimate solution (Lay, 2003). Following
strong community and local government protest, the Premier of the day, Rupert Hamer
cancelled some of the proposed inner city freeways and their road reservations (the F1,
F6, F9 and F18 which were included in the 1954 MMBW Plan, and later the F2 along
Merri Creek, see Lay, 2003, p. 202). Lay comments:
In hindsight, the plan was unacceptable in terms of cost (it would have meant a
trebling of the then-current expenditure rate on freeways), its effect on the
environment and the urban form, and its general inappropriateness for a city like
Melbourne that already had a good arterial road system’.(ibid. p.200)
Figure 7.3 of the plan shows a grid of inner city freeways that would have
destroyed much of the ring of inner suburbs from Carlton clockwise to South
Richmond (ibid. p. 201).
Construction work on both the Lower Yarra Freeway (now known as the West Gate
freeway) and the Tullamarine Freeway started in the late 1960s. The Lower Yarra
Freeway opened in April 1971 (Laybutt, undated). Work on the Eastern freeway started
soon afterwards. Its first phase was characterised at times by strong local community
opposition. Phase 1 of the Eastern freeway opened to traffic in December 1977 (Laybutt
(b), undated).
At the end of the 1980s, however, none of Melbourne’s freeways were directly
connected. To partly overcome this, construction of Melbourne’s outer ring road began in
1989, with work on its first section —the Western Ring Road — being completed in
1999. This section costs a total of $631 million, $555 million of which was provided by
the Federal government (VicRoads, nd).
The Victorian government in 1994 released a paper setting out a strategy for enhancing
Melbourne’s urban road network. One of its key recommendations was a major
expansion of Melbourne’s freeway system by substantial traffic capacity additions to
both the existing Tullamarine freeway and the South Eastern Arterial, and by ‘developing
freeway standard links between the South Eastern arterial, the West Gate freeway and the
Tullamarine Freeway (Transurban City Link Limited, 1996, p. 13)’. Collectively this
new freeway-linking project was named City Link. It is a Build-Own-Operate-Transfer
(BOOT) type of freeway initiative, incorporating fully electronic, user-pays tolling. The
total expenditure on City Link’s construction over the years 1995-96 to 1999-2000 was
valued in 1993 dollars at $1.5 billion (ACG and Cox, 1996: p. 3). Of this amount, the
Victorian government contributed around $230 million. In current dollar terms the actual
pre-opening expenditure was in excess of $1.8 billion. City Link was fully opened to
traffic in December 2000. Ownership of City Link transfers to the Victorian Government
at the end of the thirty four year Concession Deed period in 2034. In July 2007 the
Calder-Tullamarine interchange project, the aim of which was to reduce congestion at the
intersection of the Calder and Tullamarine freeways, was opened. Its construction cost
was $150 million.
4
In May 2002 construction of the Craigieburn Bypass, linking the Hume Freeway near at
Craigieburn, to the Metropolitan Ring Road at Thomastown commenced. The completed
Bypass was open to traffic in December 2005. The bypass cost $305 million and was
funded by the Australian government (VicRoads, 2006, p.46)
In May 2006 the Government announced a major freeway project to widen the section of
toll road south of the CBD between the Westgate Bridge and the Monash freeway The
Monash-City Link-West Gate upgrade is ‘a key project and will improve traffic flow and
safety along Melbourne’s busiest freeway corridor (VicRoads 2007-08 p. 5)’. . This
project termed the Monash –Westgate corridor or the M1 project was budgeted in 2006 at
$1 billion (DOI, 2006). The project is funded by the Victorian Government and
Transurban for CityLink works (VicRoads 2007-08 p. 35). By the end of fiscal year
2007-08 the total cost of this initiative was revised upward to a new figure of $1.4 billon
(VicRoads 2007-08 p. 35). At around the same time funding of $331 million was secured
from the Victorian and Australian governments for the Deer Park By Pass aimed at
providing a link between the Western ring Road and the Western freeway (VicRoads
2007-08 p. 36). This project is expected to be complete in mid 2009 VicRoads 2007-08 p.
36)..
Yet another very large freeway project undertaken in Victoria is the East Link freeway
initiative. East Link, formerly known as the Mitcham-Frankston (or Scoresby) Freeway,
is Melbourne's second fully-electronic tollway, comprising about 39kms of freeway-
standard road connecting the city's eastern and south-eastern suburbs. Construction of it
began in March 2005 and the freeway was opened for traffic at the end of June 2008.
Total construction costs for East Link were $2.5 billion (Department of Transport, 2008).
The private sector organisation, ConnectEast was chosen by the state government to
‘finance, design, construct and operate East Link. The contract was for a period of
approximately 39 years (the Allen Consulting Group., 2006). The cost benefit analysis
claims that the great majority (87.5%) of East Link’s total economic benefits of $12.9
billion would derive from the value of travel time savings (ACG, 2006). Ownership of
the East Link freeway transfers to the Victorian Government at the end of the Concession
Deed period in 2043.
In 2008, the Victorian government released a new transport vision and long-term plan for
Victoria. One of its key projects is the construction of a new tunnel ‘which will relieve
Melbourne’s dependence and reduce traffic congestion on the West Gate Bridge’ at a cost
of more than $2.5 billion .This plan also signalled its intention to ‘start the Peninsula Link
– a 25 kilometre four-lane connection between EastLink at Carrum Downs and Mount
Martha – next year in partnership with the Commonwealth’. Additionally, the Victorian
government proposed completing the ‘missing link’ between the Metropolitan Ring Road
and the Eastern Freeway at Bulleen (The Victorian Transport Plan, 2008). The cost of
this project is stated as ‘more than $6 billion’. A third major road project announced in
this 2008 transport plan is the ‘truck action plan’ aimed at removing trucks from
residential areas and improving freight access to the Port of Melbourne. Stage 1 of this
plan ‘will cost $380 million. Collectively the cost of these three new road transport
components of this ambitious transport plan amounts to in excess of $9.25 billion out of a
total investment of $38 billion ‘to significantly increase the capacity of our transport
network with more services, more often, to more places (The Victorian Transport Plan,
2008, p.3)..’
Given the magnitude of both the historical costs of additions to Melbourne’s urban road
network and the just mentioned major new investment plans, it seems both timely and highly
appropriate to compare the projected economic benefits of such major infrastructure
investments with the actual results achieved over the last fifteen or so years. The next two
sections of this paper aim to do just that by using the City Link motorway system as a case
study.
5
Projections and economic justifications for major network
projects
The Victorian government in the early 1990s commissioned two separate road network
traffic models for the whole of Melbourne’s major road networks: one by VicRoads, the
other by Veitch Lister. The objective of both network models was to compute the future
traffic volumes under two contrasting circumstances: namely (i) the Base Case in which
no change to the major road network occurs within a specified time horizon, and (ii) the
Project case in which a new road project is funded, constructed and opened for use.
The results of both of these road network traffic studies were quite similar as shown in
table 2. The simple average project case of vehicle hours of travel per day in 2001 is
1,928,400; the simple average projected savings in daily vehicle hours in 35,800; by
inference, the simple average base case daily vehicle hours is 1,964,200 (i.e. 1,928,000
plus 35800).
The Victorian government consequently commissioned two economic analysis studies of
the potential costs and benefits of City Link’s construction and operation. Both were
conducted by The Allen Consulting Group and Cox. The Allen Consulting Group will
hereafter be referred to as ACG.
Outcome type Veitch Lister VicRoads
Change in vehicle kilometres of travel if
project case were implemented (million kms
per day)
0.14 0.35
Vehicle hours travel per day (000 VHT) 1927.6 1930.2
Change in vehicle hours of travel/ day (base
case- project case) (000 VHT) -31.8 -37.9
Average travel speed urban network whole
day (project case) (km/h) 44.9 43.8
Change in average travel speed (Project
case- base case) (km/h) 1.1 1.0
Table 2: Comparison between project case forecast outcomes and base case for year 2001
Source: Extracted from ACG and Cox, 1995, Table 2.6, p. 19.
In both these cost-benefit studies, five types of economic savings were proposed and
valued: travel time savings, reduced vehicle operating costs, accident cost savings, fleet
mix savings, and off-road benefits. Of these five types of savings, travel time savings are
by far the largest, representing 63% of the total forecast future value of all benefits
accruing over the 36 years ending June 2031 (ACG and Cox, 1996: table 2.5, p. 14). The
stated assumption is that such travel time benefits will be ‘gained from the freer and
faster flow of traffic in Melbourne’ should City Link be constructed and subsequently
used (ACG and Cox, 1996: 9). In particular, City Link ‘will have a dramatic effect on the
efficiency of travel movements within the central core of Melbourne’ (ACG and Cox,
1995: 4).
The 1996 benefit-cost study of City Link presents a comparative forecast drawn on the
modelling by Veitch Lister Pty Ltd for use by the Melbourne City Link Authority
(MCLA) in evaluating the project tenders for City Link (ACG and Cox, 1995: 17). It is
presented here as table 3. It is important to note that the projections set down in table 3
are made on the basis that Transurban had actually been awarded the BOOT tender and
that ‘a number of other projects are deemed to have been completed by 2011 (e.g. the F2
freeway, the Scoresby Freeway – now East Link, and the Mornington Peninsula Freeway
(ACG and Cox, 1996: 2). The two latter have now been completed.
6
Road
type Vehicle type Base case
:City Link
not built
(000s daily
vehicle
hours)
Project case:
City Link
built
(000s daily
vehicle
hours)
CL minus
base case:
(000s
daily
vehicle
hours)
Freeways Private 229,000 236,000 7,000
Commercial
97,000 95,000 -2,000
Total Freeway DVH 326,000 331 5
Arterial Private 1348 1320 -28
Commercial 453 446 -7
Total arterial DVH 1801 1766 -35
Grand total DVH 2127 2097 -30
Table 3:
Projected daily vehicle hour time savings from City Link: year 2011 (Source: ACG & Cox
1996, Table 2.1, p. 11estimates by Veitch Lister Pty Ltd for MCLA)
In sum, table 3 indicates that the construction of City Link was projected to reduce total
daily vehicle hours (DVH) on Melbourne’s urban road network by 30,000 hours — or
around 1.4% — per day, as compared with the ‘base case’ if the freeway system were not
expanded.
Mathematically, the specific value of three variables must be determined before daily
vehicle hours (DVH) can be computed:
(i) number of trips per day (T)
(ii) average distance per trip (kms) (D); and
(iii)average speed per hour (kms/ hr). (S)
The product of variables (i) and (ii) equals vehicle kilometres travelled (VKT). VKT
divided by speed per hour (S) = DVH
These variables are now considered:
Number of trips per day
The projected number of trips per day on either freeways or arterial roads is not explicitly
stated. As noted in table 3 ACG and Cox (1996) contend that the total number of
vehicular trips taken on Melbourne’s freeways will increase by around 1.5%, presumably
due to reduced freeway congestion resulting from a significant expansion of total freeway
carrying capacity and a corresponding increase in the attractiveness of freeway travel to
Melbourne’s motorists. The reduction in total daily vehicle hours travelled on
Melbourne’s arterial roads of just under 2% detailed in Table 3 presumably is the result
of both the just mentioned growth of trips per day on Melbourne’s expanded freeways
that in turn would reduce the number of trips per day on the arterial roads and thus enable
faster average travel speeds overall. The only explicit comment offered in the 1996 cost-
benefit analysis of City Link was that annual growth in traffic volume is assumed to be 2
% per year (ACG and Cox, 1996:13).
Average distance per trip
The second variable determining daily vehicle hours is average distance per trip. Again,
no specific data on this variable were disclosed in either cost-benefit study. One is
probably entitled to assume, however, that the expansion of Melbourne’s freeway system
that the City Link project would provide would result in a reduction in distance per trip
for motorists who moved from travelling on arterial roads to travelling on the expanded
freeway network, since freeways are generally more direct in route than arterial roads.
7
Average speed per trip
As already noted, the third variable needed to compute daily vehicle hours is average
speed per trip taken. The achievement of this projected total reduction in travel times
across Melbourne’s road network was claimed to result directly from a corresponding
projected increase in average speed, especially within inner Melbourne.
Most of the travel time savings (74 per cent) are located in an 8kms x 8 kms grid close to
the CBD even though this area has only 7 per cent of the total travel in the metropolitan
area. The average travel speed in this inner city area will rise significantly from 29.7 km /
hr to 36.1 km/hr compared to a minor increase of 0.3 km/hr outside of this grid (ACG and
Cox, 1995: 19). Value of travel time savings
The next value proposed in the City Link benefit cost study was the most important one:
the hourly value of travel time savings. ACG and Cox (1996) argued for the use of a
composite hourly rate of $19.15. This composite derives from the addition of two thirds
of an hourly rate of $21.50 for the value of time saved in inner Melbourne, and one third
of an hourly rate of $14.42. This valuation was based on analysis by Thorensen,
Thorensen and Taylor (as cited in ACG and Cox, 1995: 20). The justification for the use
of a considerably higher dollar value than that used by other analysts for travel time
saved was that the bulk of the (projected time) ‘savings are located in the inner
Melbourne urban area where there is more commercial travel and which should lead to a
significantly higher value of time than for an average vehicle distribution around
Melbourne (ACG and Cox, 1995: 21)’.
Annual average time savings
Two other metrics are needed in order to compute annual travel time savings. Without
substantive justification, the assumption was made that 330 days a year was appropriate
in order to ‘aggregate annual values’ (ACG, 1995: 21). The final metric required is the
average annual rate of growth of travel time savings over the economic life of City Link.
Without supporting justification it was also stated that ‘time benefits grow at 4.25 per
cent per year (ACG, 1996: 13).’ Such a projected annual growth in travel time savings is
used even though the assumed annual growth in traffic volume is less than half of that
percentage: i.e. 2 % per year (ACG, 1996:13).
The resultant projected annual dollar values of time benefits computed (see ACG 2006
Table 2.5, p. 14) have been converted into equivalent daily vehicle hours saved by
applying the following formula. DVHs = 1000*($b/ $ts /n)
where:
DVHs = Implicit daily vehicle hours saved (‘000 hrs)
$b = projected value of travel time savings (million $)
$ts = hourly value of travel time saved
n = number of days/ year savings achieved.
Table 4 presents the ACG & Cox (1996) forecast of DVH that would be saved per year
form 1997-08 to 2011-12. It is most surprising however that time savings were projected
for the years 1997-98 through to 1999-2000 given that City Link’s construction would
not be complete until the middle part of the year 2000. The consequent disruption of
traffic flows especially on the Tullamarine freeway was specifically noted:.
Since mid 1996, construction works on City Link, coupled with the staged opening of the
Western Ring road, (in June 1997) in particular were negatively affecting travel speed
trends on the Tullamarine freeway (Vic Roads, 2001).
.
8
Comparison and analysis of projected performance against actual
performance
Table 5 presents the main projections that underpinned the quantification of the value of
the annual travel time savings presented in table 4, and a comparison with publicised
actual results over the years 1994-5 to 2006-7.
As shown in table 5 there was a negative variance of 2.3 kms / hour (or 6.4%) between
the projected and the actual average travel speeds on all roads in inner Melbourne in
2001. Analysing traffic patterns on a whole of day basis in major metropolitan centres
like Melbourne however obscures the quite variable nature of road usage by hours across
a typical weekday. Figure 2 shows clearly this time-of-day road use variability. It
highlights the clearly bi-modal nature of ‘traffic distribution’ in Melbourne for both
freeways and for arterial roads. On Melbourne’s freeways morning traffic volumes spike
sharply from around 5 am to about 7.30 am. After that, traffic volumes drop until around
11 am and then rise gradually until about 2.30 pm. The afternoon peak commences
around 3.30 pm and ends around 6.30 pm. As noted in Figure 2, traffic volumes per
fifteen minute intervals fall very sharply after that. The other notable feature depicted in
Figure 2 is that the morning peak for freeways both starts and finishes earlier than that for
arterial roads.
Table 5:
Basic propositions underpinning projected City Link travel time savings
Propositions Actual
Variable Year Specific
value
projected
Actual value
reported 6
1. Average travel speed inner
Melbourne; all roads whole of
day (km /hr)1
2001 36.1 33.8
2. Average travel speed across
urban road network (km/ hr) 2 2001 44.3 42.6
(2000-01)
3. Total daily vehicle hours of
travel urban road network (000
hrs) 3
2001 1929.8 1991.8
(1999-00)
2087.2
(2000-01)
4. Total daily vehicle hours of
travel urban road network (000
hrs) 4
2011 2097.0 2157.2
(2006-7)
5. Melbourne’s freeways total
daily vehicle hours (000s)4 2011 331.0 244.7
(2006-07 )
6. Melbourne’s arterial roads
total daily vehicle hours (000s)4 2011 1766.0 1866.8
(2006-07)
7. Freeway travel’s share of
total vehicle kms of across
urban network 4
2001 (i) 17.3%
(ii) 20.3% 21.9%
8. Total cumulative change in
daily vehicle hours (DVH) 5 1997-8 to
2006-07 -162 328.7
Sources:
1. ACG and Cox, 1995, p. 19.
2. VicRoads as cited in ACG and Cox, 1995, p. 19. Note 44.3 is simple average of
43.8 km projected by Veitch Lister and 42.8 kms projected by VicRoads.
3. VicRoads as cited in ACG and Cox, 1995, p. 19. Note 1927.6 is simple average of
1959.4 projected by Veitch Lister and 1930.2 projected by VicRoads
4. Veitch Lister as cited in ACG and Cox, 1996, p. 11. 17.3% projected by Veitch
Lister; 20.3% projected by VicRoads.
5. ACG and Cox (1996)
6. All actual results obtained from VicRoads (2009) Annual Traffic System
Performance Monitoring; longitudinal data from 1994/5 to 2006/7 provided to
author by Vic Roads April 2009.
Figure 2: Variability of road usage over a 24 hour cycle: Typical week day
(Source: VicRoads (2008): Traffic System Performance Monitoring 2006-
07, p. 12)
The average speeds for the freeways during both the morning and the evening peaks in
the inner Melbourne region over the year 1994-95 to 2006-07 have increased: from 46.8
km /hr to 58. 8 km/ hour for the morning peak and from 67.8 km/ hour to 73.5 km/ hour
for the afternoon peak (Vic Roads, 2009). However, for the period 2001-02 to 2006-07,
the average freeway speed per hour in the inner Melbourne region has dropped markedly,
both in the morning peak (from 67. 4 km/ hr to 58.8 km/ hr) and in the evening peak
(from 80.2 km/ hr to 73.5 km /hr). Given the fact that City Link was fully operational in
the second half of 2000-01, and the very substantial economic investment that it
represents, one must be concerned about this quite substantial drop in average speed per
hour on the busiest usage periods of the day. The average speeds on Melbourne’s
freeways for the entire urban network overall have also dropped in the years following
City Link’s full opening. Figure 3 illustrates this drop in comparison to a steady increase
in total vehicle kilometres travelled per year across Melbourne’s urban road network.
One possible explanation for this overall reduction in average speed on Melbourne’s
freeways is that the morning peak period is now 30 minutes longer than in 2001/02 and
the afternoon peak is up to 30 minutes longer (Vic Roads, 2007). Another plausible
explanation is that freeway travel in the evenings as a proportion of the whole day is
lower than for arterial roads (Vic Roads 2006)’. In other words the bulk of the increased
total freeway VKT noted in Figure 3 is occurring at times when the freeway’s capacity to
absorb extra traffic (whilst allowing vehicles to maintain higher average speeds) is at its
lowest.
The average speeds on Melbourne’s inner arterial roads during both the morning and the
afternoon peak periods have also fallen over the period 1994-5 to 2006-7. For instance
the average speed in the morning peak on inner Melbourne’s divided arterials fell from
34.3 km /hour in 1994-5 to 29.5 km/ hour in 2006-07 (Vic Roads, 2009). The same trend
applies in the evening peak that has experienced a fall in average speed per hour from
34.9 kms to 30.9 kms over the same period. Admittedly the year by year pattern is
volatile with some years seeing an increase (1996-7 and 2002-03 the most notable). A
similar reduction in average speeds in both the morning and evening peaks is observable
for inner Melbourne’s undivided arterial roads.
9
0
5000
10000
15000
20000
25000
1994-951996-97 1998-99 2000-012002-03 2004-052006-07
Year
Total VKT (Million kms)
55.0
60.0
65.0
70.0
75.0
80.0
85.0
90.0
Av. speed all day (kms/hr)
Total VKT Av. Speed (kms/ hr)
Figure 3: Whole of day average speed / kilometre on Melbourne’s freeways
compared to total kilometres travelled on freeways
In summary, average travel speeds in inner Melbourne post the opening of City Link
have reduced in both the morning and evening peaks. Even more concerning is the fact
that average speeds across the whole day for both freeways and all types of arterial roads
in the inner Melbourne region have all similarly dropped over the years 2001-02 to 2006-
07. Figure 4 represents the simple average speed per kilometre for inner urban arterial
roads over the tears 1994-5 to 2006-7 plotted against total VKT on these roads.
15000
15500
16000
16500
17000
17500
18000
1994-95 1996-97 1998-99 2000-01 2002-03 2004-05 2006-07
Year
Total VKT (million kms)
26
27
28
29
30
31
32
33
Av. speed (kms/ hr)
Total VKT Simple average speed (whole day)
Figure 4: Melbourne's inner arterials: Total VKT and simple average speed
all day
The second variable specified in Table 5 is the average travel speeds for the whole of day
across the whole urban network. In 2001, the prediction was for an average speed of 44.3
kms/ hour. The actual average speed for the year 2000-01 was 42.6 kms/ hour. Over the
years 1994-95 to 2006-07 this average speed per hour across whole of day for the urban
road network has dropped by one kilometre per hour. Over the years 2001-02 to 2006-07
it has fallen by 0.6 kilometres per hour from 41.4 in 2001-2 to 40.8 in 2006-7.
The claim that traffic will flow more freely after City Link was completed has been tested
by examining the reported changes in traffic time variability across the urban road
10
network. Figure 5 presents the reported data for the years 1996-97 to 2006-07 for the
urban road network monitored by Vic Roads. Based on these data the projected increase
in flow of traffic does not appear to have been achieved.
Variability of travel time Melbourne urban road network
10
12
14
16
18
20
22
24
26
28
1996-97 1997-98 1998-99 1999-00 2000-01 2001-02 2002- 03 2003-04 2004-05 2005-06 2006-07
All roads, all day Al l r ao ds, non peak Al l r aods A M peak A ll r aods, PM peak
Figure 5: Trend in variability of travel time by time period 1996-7 to 2006-
7
(Source: Vic Roads Traffic System Performance Monitoring reports: Series 1: AM
peak; Series 2: PM peak Series 3: off peak Series 4: all day.)
Another piece of evidence disputing the claim that City Link would result in freer
flowing traffic is presented as table 6. It indicates that congestion (as measured by traffic
delay in minutes per kilometre travelled) has not decreased over the nine years to 2005-
06. VicRoads (2008) itself notes that ‘the the aim is for the rate of increase in traffic
delay to be less than the rate of increase in road travel.(…) On average over the past five
years this aim has not been met, with delay growing at a higher rate than road travel
(VicRoads, 2007—08 p. 43)’.
Table 6: Traffic Levels and Congestion (At urban monitored locations, all
times)
Year
Vehicle
kilometres
travelled
Traffic delay in
minutes per km
travelled1
1997-98 78
,
318 0.53
1998-99 80,785 0.50
1999
-
00
82,803
0.50
2000
-
01
85,030
0.51
2001
-
02
86,412
0.55
2002
-
03
86,460
0.53
2003
-
04
88,301
0.52
2004
-
05
88,688
0.57
2005
-
06
88,041
0.
53
Source: VicRoads, Annual Report 2006-06, page 57.
Given the observed negative disparity between projected and actual speed in both the
inner Melbourne area, and in the urban road network overall, urban network travel time
savings would be possible only if the actual percentage traffic volume using Melbourne’s
inner roads fell during the relevant time period and at the same time the average speed on
Melbourne’s outer roads increased. As just noted, the average speeds have not increased.
Between 1994-5 and 2006-07, the percentage of total vehicle kilometres travelled (VKT)
in the urban Melbourne road network represented by inner Melbourne traffic has actually
fallen from 27.5% (1994-95) down to 23.9% (2006-07). This decline has been a gradual
11
and relatively steady one. Moreover the simple average speed achieved on Melbourne’s
outer urban roads for the whole day has actually declined from around 44.1kms/hr in
1995 to 40.8 kms /hour in 2006-7, as illustrated in figure 6.
45000
46000
47000
48000
49000
50000
51000
52000
53000
54000
1994-95 1996-97 1998-99 2000-01 2002-03 2004-05 2006-07
Year
Total VKT (million kms
37.0
38.0
39.0
40.0
41.0
42.0
43.0
44.0
45.0
Av. speed (kms/ hr)
Outer urban arterials Total VKT
Outer arerial Simple average speed all day
Figure 6: Melbourne's outer arterials: Total VKT and simple average speed
all day
The third to seventh variables specified in table 5 all involve total daily vehicle hours
travelled. Table 7 presents actual data for the period 1994- 95 to 2006-07. This table
shows that the projected total DVH with City Link in year 2001 of 2,097,000 hours (see
Table 5) was actually exceeded in the year 2004-05 and was above the City Link forecast
level in both 2005-06 and 2006-07.
The projected volume of freeway traffic of 331,000 DVH (table 2) by 2011 looks
unlikely based on the reduced speed of freeway traffic volumes since 2003-04 and a total
freeway volume of 244700 DVH in 2006-07 (Table 7). Despite this, the growing
popularity of Melbourne’s freeways is clearly evident. Indeed the actual share of total
kilometres travelled within Melbourne’s’ urban road network in 2006-7 accounted for by
freeways has exceeded the projections by Veitch Lister and by Vic Roads (see table 5).
The last comparison presented in Table 5 is between the projected amount of travel time
saved — as measured by reduced daily vehicle hours — through the introduction of
additional freeway capacity in Melbourne and the actual annual increment to daily
vehicle hours recorded and reported by Vic Roads. Table 8 and Figure 6 address this
comparison
Table 7: Actual Daily vehicle hours (dvh) Melbourne urban road network
Year
Est. total
daily
vehicle
hours all
freeways
(000hrs)
Non
freeway
urban
network
est. total
dvh (000
hrs)
Est. total
daily
vehicle
hours
(000 hrs)
% of total
daily vehic
hours
travelled o
n
freeways
Est. total
vehicle km
travelled
1994-95 123.3 1674.6 1797.9 12.8% 75153
1995-96 130.0 1688.6 1818.7 13.3% 76384
1996-97 145.1 1663.1 1808.2 15.0% 77030
1997-98 163.7 1705.5 1869.2 16.7% 78318
12
13
1998-99 174.4 1700.0 1874.4 17.1% 80785
1999-00 197.3 1729.7 1927.0 18.55% 82862
2000-01 196.5 1795.3 1991.8 19.24% 84850
2001-02 210.2 1877.1 2087.2 19.3% 86412
2002-03 227.1 1842.3 2069.4 20.4% 86913
2003-04 236.9 1848.2 2085.1 20.8% 88408
2004-05 243.2 1923.0 2166.2 21.1% 88598
2005-06 243.5 1893.4 2136.9 21.5% 88041
2006-07 244.7 1912.5 2157.2 21.9% 88015
Source: 1994-95 to 2006-07 data supplied by VicRoads electronically on 8 April
2009.
The lack of travel time savings that were projected and used to justify the economic
benefits that would result from both City Link and more recently East Link is clearly
evident in both table 8 and in figure 7.
Table 8:
Comparison of actual and projected changes in Daily vehicle hours
Melbourne urban road network 1997-8 to 2006-07
Year
Est. total
DVH Urban
Melbourne
network (000
hrs)
Annual
change in
total DVH
(000
Imputed
project case
DVH -base
case DVH
(000 hours)
Annual
change
Actual DVH
- Base case
change DVH
1996-97 1808.2
1997-98 1869.2 61.0 -1.9 62.9
1998-99 1874.4 5.2 -1.9 7.1
1999-2000 1927.0 52.6 -9.4 62.0
2000-01 1991.8 64.8 -18.7 83.5
2001-02 2087.2 95.4 -19.5 114.9
2002-03 2069.4 -17.8 -20.3 2.5
2003-04 2085.1 15.7 -21.2 36.9
2004-05 2166.2 81.1 -22.1 103.2
2005-06 2136.9 -29.3 -23.0 -6.3
2006-07 2157.2 20.3 -24.0 44.3
Total 349.0 -162.0 511
-150.0
-100.0
-50.0
0.0
50.0
100.0
150.0
13579111315171921232527293133
Year
DVH saved /lost (000 hrs)
2006-7
Actual change in DVH (& required to achieve total
p
rojected travel time savings NPV
Projected DVH saved (ACG &
Cox 1996)
Figure 7 Melbourne urban road network: projected and actual travel time
savings: 1997-8 to 2006-7
The Net present Value (in 1993 dollars) of time savings projected for City Link over
thirty six years was $1.4851 billion (ACG and Cox, 1996). Of this total amount, $528.1
million were projected to be saved during the years 1997-8 to 2006-07. With the same
discount rate of 8% the NPV of the extra travel time in daily vehicle hour equivalents
over the years 1997-98 to 2006-07 shown in table 8 is -$349.4 million (in 1993 dollars).
Given this net cost (dissaving) to the Victorian community, computations reveal that in
order for the total projected travel time savings NPV of $1.485 billion to be achieved, the
NPV of travel time savings on Melbourne’s’ urban road network from 2007-08 to 2030-
31 inclusive would have to amount to $1.834.4 billion., as opposed to the projected NPV
of $957.1 million.
Given the actual shortfall in travel time savings already noted future travel time savings
would have to increase in value by more than 64% per year faster than they were forecast
to grow post 2011 by ACG and Cox (1996). As noted earlier, time savings were forecast
by ACG and Cox (1996) to grow by 4.5 per year. They would need to grow year-on-year
consistently at almost double this rate for the total NPV of time savings to be achieved.
Based on the above analysis of actual urban network performance such an outcome
seems highly unlikely, at least in the absence of an extremely benevolent deus ex
machina.
Conclusions
Several conclusions can be drawn from this empirical analysis:
(i) the actual volume of traffic on Melbourne’s urban road network in the
years 2000-1 to 2006-7 , measured on both total vehicle kilometres and
daily vehicle hours, exceeds the volume forecast in the early nineteen nineties;
(ii) the rate of growth in daily total vehicle kilometres travelled slowed to
0.6% per year during the years 2001-02 to 2006-07, compared to 2.1%
per year from 1994-5 to 200-01;
14
(iii) whole of day average travel speeds per hour have increased on Melbourne’s inner
freeways; however, the average whole of day speed on Melbourne’s freeways overall has
stayed at around the same level (78+ kms /hour) apart from 2000-01 (83.5 kms/ hour) and
2001-02 (79.5 kms’ hour). However, over the years 2001—02 to 2006-07, the average
kilometre per hour speed on Melbourne’s inner freeways in the morning peak has
dropped from 67.4 kms/ hr to 58.8 kms/ hr; in the afternoon period the fall has been from
80.2 kms/ hr to 73.5 kms /hr.
15
(iv) the whole of day average speed per hour across the monitored urban road network in
2006-7 was 1 kilometre per hour slower than in 1994-5;
(v) reductions in daily vehicle hours travelled on Melbourne’s arterial road network have
occurred; but the corresponding increase in average speed per hour has not: for instance,
the average whole of day speed on inner divided arterials has dropped from 39.8 kms/
hour in 2000-1 to 34 kms/ hour in 2006-7; and for outer area undivided arterials it has
dropped from 41.4 kms/ hour in 200-1 to 39.5 kms/hr in 2006-7.
(vi) the average speed in kilometres per hour in both the morning and evening peak
periods for the whole monitored urban network in the most recent year for which data are
available — the year ended June 30 2007 — is the lowest it has been since 1994-95.
In sum, these results suggest that the core of travel times savings benefits, which is an
increase in average travel speeds, has not eventuated in Melbourne’s urban road network
during the years under review. Indeed, based on the evidence presented and analysed in
this paper, one could be led to the conclusion that investments in Melbourne’s urban road
network have resulted in more time being used by Melbourne’s motorists rather than less
time. On the basis of the assumptions conventionally used to justify road building , major
road infrastructure initiatives have resulted in net economic disbenefits.
We may acknowledge that people derive benefit from freeways, as Metz (2008) points
out, in opening up new opportunities for interaction further afield. He writes, ‘My
argument is that the benefits of new transport infrastructure are taken by travelers in the
form of greater access, not time saving’ (Metz, 2008 p.31). But, as he continues, ‘The
transport economists have led us badly astray’. A more dispersed, car dependent city
benefits some, but also has serious costs in terms of greenhouse emissions and oil
dependence.
This conclusion is most concerning, given the very substantial amounts of public funds
that have been directed at Melbourne’s urban road network over the last several decades,
coupled with the dominance of claimed travel time savings in the economic cost-benefit
justifications that have preceded the approval of such very large scale additions to
Melbourne’s freeway stock. They are even more troubling when one acknowledges the
comparatively higher length of freeway per capita that Melbourne enjoyed over
numerous comparable cities both in Australia and overseas (Kenworthy and Newman
2002) , even before East Link and the other planned expansions to our freeways came or
will come on stream.
This analysis clearly raises very real doubts about the validity of using the projected
value of travel time savings as the basis for estimating the social economic benefits of
building new road projects, at least within established and complex urban road networks.
Even if this validity were accepted, there is a very strong need for far more rigorous and
empirically based methods to be used when projecting the quantum of travel time most
likely to result from the road project being evaluated at the time.
In closing, perhaps the German word schlimmbesserung — meaning an improvement that
makes things worse — is an apt descriptor for the massive program of new road
construction that has marked Melbourne’s ‘solution’ to its transport challenges over the
last several decades.
References
Anon (2008): ‘Melbourne freeways upgrade to cost $1.4 billion”, Drive Fairfax Digital, May 1.
http://www.drive.com.au/Editorial/ArticleDetail.aspx?ArticleID=52057&vf=26
accessed 28 April 2009
Davidson, K. (2008) ‘Wrong turn - freeway funds a waste’, The Age, May 8.
http://business.theage.com.au/business/wrong-turn--freeway-funds-a-waste-20080504-2avj.html
accessed 29/04/09.
Department of Infrastructure (2006) Annual Report 2005-06, Melbourne, Victorian Government.
Department of Transport Annual Report 2007-08, Melbourne, Victorian Government 29 October
2008.
16
Ironmonger, D. 2008, A System of Time Accounts for Melbourne, Second Edition, Report
Commissioned by The Department of Transport, Victoria, Australia, Households Research Unit,
The University of Melbourne, Melbourne, Victoria. Australia.
Kenworthy, J. & P. Newman, P. “Melbourne in an International Comparison of Urban Transport
Systems: Marvellous or Mediocre”, Institute for Sustainability and Technology Policy, Murdoch
University, 2002. p.19
Lay, M. and Daley, K. 2002 ‘The Melbourne City Link Project, Transport Policy, 9, pp. 261-67.
Lay, M.G. (2003) Melbourne Miles Melbourne: Australian Scholarly Publishing
Laybutt, S. ‘History of the M3 Eastern Freeway’
(http://mrv.ozroads.net.au/SRNS/M%20Routes/M3%20Eastern/history.htm) 25/04/09.
Melbourne City Link Authority, 2001, Annual Report, Victorian Government, Melbourne.
Melbourne Transport Committee (undated): Melbourne’s Transport Task – an overview. Open,
working document.
Metz, D (2008) The Limits to Travel, How far will you go? London: Earthscan
Minister for Transport (2002) ‘BRACKS GOVERNMENT ANNOUNCES MITCHAM-
FRANKSTON FREEWAY PROJECT Media Release, September 23.
http://www.dpc.vic.gov.au/domino/Web_Notes/MediaRelArc02.nsf/fce5683f42af57f04a256b3600
1bd4df/d00505afe9c35a32ca256c3d007c472c?OpenDocument accessed 28 April 2009.
Parsons Brinkerhoff (2008): “TCI’
http://www.pb.com.au/PBAU/Projects/Tullamarine%E2%80%93Calder+Interchange/ACCESED
26/04/09.
Public Transport Users Association (2008) ‘Common Urban Myths About Transport’
Salt, B. (2008): The EastLink Imperative, report commissioned by ConnectEast Group, July.
The Allen Consulting Group and Cox, J. 1995 The Economic Impact of City Link, Consultancy
Report to the Melbourne City Link Authority, May, Melbourne.
The Allen Consulting Group and Cox, 1996, The Economic Impact of City Link: Transurban
Project Consultancy Report to the Melbourne City Link Authority, April, Melbourne.
The Allen Consulting Group (2006): The Economic Impacts of East Link Report to the Southern
and Eastern Integrated Transport Authority, 21 August.
Transurban City Link Limited, 1996, The Melbourne City Link Prospectus, 1996
Victorian Department of Transport, 1991: Central Area Transport Strategy, Melbourne, Victoria,
AU, March.
VicRoads (2001): Traffic System Performance Monitoring 1999-2000, Melbourne.
VicRoads 2000-2001 Traffic System Performance Monitoring Report.
VicRoads 2001 -2002 Traffic System Performance Monitoring Report.
VicRoads 2002-2003 Traffic System Performance Monitoring Report.
VicRoads (2006) 2005-06 Annual Report, Melbourne.
VicRoads (2008): Traffic System Performance Monitoring 2006-07, Melbourne, Vic, AU.
VicRoads (nd). ‘History of VicRoads’
http://www.vicroads.vic.gov.au/Home/AboutVicRoads/HistoryOfVicRoads1.htm 25/04/09
VicRoads: History of Western Ring Road
‘(http://www.vicroads.vic.gov.au/Home/AboutVicRoads/HistoryOfVicRoads1.htm 25/04/09)
VicRoads (2007) VicRoads Annual Report 2007-08
VicRoads (2009) Annual Traffic System Performance Monitoring; longitudinal data from 1994/5
to 2006/7 provided to author by Vic Roads April 2009.
VicRoads (2009b): ‘Craigieburn Bypass Community Opening Event
http://www.vicroads.vic.gov.au/Home/RoadsAndProjects/RoadProjects/Expired+Project+Pages/C
raigieburn+Bypass+Project.htm, accessed 23 August 2009.
Victorian Government, 2008, The Victorian Transport Plan, Treasury Place, Melbourne, Victoria
Wikipedia: Western Ring Road (http://en.wikipedia.org/wiki/Western_Ring_Road 25/04/09)
Williams, P. J. 1992, Swanston Street Closure Travel Time Surveys RACV, Melbourne, Victoria,
AU.
