The current franchising arrangements for Melbourne’s trains and trams have been extensively criticised by the PTUA and others for neglecting the public interest. Such criticism, particularly in media forums, has of necessity not always been comprehensively articulated, and has led some to misconstrue the underlying facts. This appendix, accordingly, is devoted to presenting some of the most common misunderstandings and supplying the necessary factual details.
- Subsidies are only higher because we’ve bought new rolling stock.
- The Auditor-General’s report said franchising was value for money!
- Thirty-five people can’t possibly do the work of three hundred!
- But look at all the responsibilities the Director of Public Transport already has: how is that any different from what the PTUA proposes?
- What about the Director of Public Transport’s new powers in land use planning?
- Why all the fuss over petty distinctions between franchise contracts and other kinds of contract?
- Zurich’s public transport succeeds because they’re good at doing timetables, not because of some quirk of bureaucratic organisation.
- In Zurich the train operator writes the timetables, not the ZVV!
- Zurich’s population is only one-tenth that of Melbourne!
- There are freeways being built in both Vancouver and Perth!
- Overseas experts think our public transport system is wonderful!
- What hope is there for public control when The Met was so hopeless?
- Reliability and punctuality have improved since privatisation.
- The ‘privatisation debate’ is a distraction that is holding us back from pursuing real improvements to public transport.
The Auditor-General’s report shows this to be untrue. According to Figure 2E of the report, the privatised system cost approximately $1.2 billion more to operate up to financial year 2006 than maintaining public operation at the 1999 level. Figure 2D, meanwhile, shows the total cost of operating Melbourne’s trains and trams, on both an actual and “normalised” basis. The difference between the actual and normalised figures represents an adjustment for “constant investment in rolling stock” and in 1998–99 is approximately $100 million. Since there was very little rolling stock investment in 1998–99, it follows that the cost of the new rolling stock is equivalent to $100 million each year, which represents $700 million up to financial year 2006. This leaves an unexplained excess of $500 million, or 25% more than would be expected under a continuation of public operation. (All figures are in 2005 dollars.)
It is also questionable whether the new rolling stock is actually worth what the government has paid for it. The brake problems with the Siemens trains are now well known, and it was revealed last year that the Combino trams are subject to a recall from the manufacturer due to structural faults. The train and tram fleets have also been rendered more complex and hard to manage by the fact that the two original operators for each mode each bought different vehicles. The fact that some of the suppliers are related companies of Veolia and Transdev is also of concern.
Allsop (2007) presented a calculation purporting to show that real subsidies are only $18 million higher in 2007 than in 1999, and this additional figure is easily explained as the cost of additional services. However, Allsop includes in the 2007 calculation a $75 million figure for inflation, ignoring the fact that the figures in the Auditor-General’s report are already adjusted for inflation and presented in 2004-05 real dollars (as the heading of Figure 2E makes clear). Allsop’s calculation actually confirms that the subsidy is at least $75 million higher than the cost of operation would suggest.
These same erroneous figures were repeated in a ‘value assessment’ of franchising produced by the Department of Infrastructure in August 2007. This report was prefaced by a ‘statement of review’ by Deloitte Touche Tohmatsu that may have given some readers the impression that the report was audited. However, the statement from Deloittes is not an audit opinion and makes clear that Deloittes are not in any way certifying the conclusions of the report (Department of Infrastructure 2007).
What the Auditor-General actually found was that the cost of the second privatisation was in keeping with ‘public sector benchmarks’ established by the Department, and that the process for establishing these benchmarks was appropriately documented. It did not address the content of the ‘public sector benchmarks’ themselves.
In other words, the Auditor-General’s report was devoted almost entirely to ‘process’ issues rather than the accounting principles used in the value assessment. The actual questions posed in the body of the report are:
- Did the responsible agencies provide effective advice to government?
- Did DoI allocate risks effectively?
- Were public sector benchmarks developed and used effectively during the negotiation process?
- Did DoI effectively use other negotiation strategies?
- Did DoI observe probity requirements?
- Did DoI build in adequate performance monitoring arrangements?
It is entirely possible to answer all these questions in the affirmative, even if the price benchmarks themselves were wildly inaccurate and based on creative accounting.
In the Kennett era, when similar ‘public sector benchmarks’ were employed as part of the new compulsory competitive tendering process for local councils, the benchmark methodology was criticised by many as artificially loading up the public sector alternative with invented ‘costs’ for the benefit of private-sector bidders. A similar form of accounting device appears to have been used by the Department of Infrastructure in this case, holding the public sector responsible for ‘capital charges’ on 100 years’ worth of urban rail infrastructure and rolling stock that the private sector somehow avoids.
Supporters of the current arrangements (Allsop 2007) have also tried to deflect attention from the payment figures in the Auditor-General’s Figure 2E by calling attention instead to the ‘total cost of operation’ in Figure 2D. We have little argument with the view that the cost of operating the system has changed little since 1999—after all there have been only minor improvements in service since then. The issue is whether the private operators are being paid more than the old PTC to run a system that costs about the same to run. This is what makes Figure 2E so important.
The ZVV has an organisation chart on its website in which all its staff, including the receptionist, are individually named. There are 35 people in total, who between them coordinate all public transport in the Canton of Zurich. They do not write all the timetables, but they do determine the frameworks for producing timetables and work closely with partner organisations to produce them.
The Office of the Director of Public Transport includes some 340 staff, and Metlink several more, whose collective role is a subset of that of the ZVV: that is, they are notionally responsible for collecting and distributing fare revenue, overseeing infrastructure development, coordinating timetables, and strategic planning in general. (We exclude the front-line Metlink staff whose function is selling tickets.) Their role is only a subset of the ZVV’s because, as we have explained, the ZVV takes a much more interventionist role in tactical planning.
The reason for the great disparity between the 35 staff at the ZVV and the 350-odd in the Victorian bureaucracy is that most of the latter are not actually doing any planning: they are occupied in contract administration, enforcing compliance of the private operators with the terms of their 400-page franchise agreements. The sheer complexity of the agreements makes this a task capable of occupying a very large number of people. But as explained above, it is a task peculiar to the franchising model of transport governance. In the Transport Community model as applied in Zurich, operators largely administer their own contracts (ZVV 2007). While planning agencies do arrange for inspectors to check vehicles for cleanliness, much of the ‘compliance’ becomes routine when the operator’s job is simply to follow the given timetable and routes.
Consider for example just one provision of the franchise contracts: the calculation of payments. According to the 29-page Schedule 14 of the Connex contract (for example) there are 18 types of payment that occur each month or quarter:
- Fixed monthly franchise sum
- Force Majeure adjustments
- Concession top up payment
- Special Event balancing payment
- Rolling stock adjustment
- Revenue risk sharing payment
- Profit sharing payment
- Service quality incentive payment
- Service growth incentive payment
- New ticketing revenue guarantee payment
- Employee leave entitlements payment
- Employee long service leave entitlements payment
- Bayside Committed Rolling Stock Maintenance employee entitlements payment
- Employee transfer adjustment payment
- Redundancy reimbursement payment
- Access charge adjustment payment
- Inventory value adjustment payment
- Fare change adjustment payment
Most of these payments are made by the Director of Public Transport to Connex, but some payments travel in the opposite direction. Each is calculated according to a complicated formula that requires all manner of input quantities to be monitored. Note that the penalties for late running and cancellations are not included here: they have their own section (Schedule 7) comprising 47 pages of definitions, formulae and lists of timing points. (We’ve also not bothered to list the pages of special payments related to the Commonwealth Games.) And the Yarra Trams contract is equally Byzantine, showing that there is nothing especially mode-specific about all this complexity.
We can compare this with the contracts issued by the Greater Copenhagen Authority, which subcontracts bus services on a competitively tendered, fee-for-service basis. (We choose this example mainly because the authority publishes English versions of its contracts.) These contracts devote just 5 pages to describing how payments are adjusted to reflect changes in input costs, and just 6 pages to describing the ‘quality control’ model and the calculation of associated financial bonuses and setoffs (HUR 2003). Payments to the operator in the Copenhagen model have just 5 components in total (including penalties):
- Indexed service fee
- Adjustment for ticket sales
- Adjustment for agreed service variations
- Quality control setoffs and bonuses
- Fines for late running or cancellations
Worth noting is that the Copenhagen quality control model includes passenger comfort factors such as ventilation and temperature control in vehicles, which our franchise contracts do not address in all their 400 pages, other than through general maintenance obligations and by requiring the operators to conduct customer satisfaction surveys (which shifts the onus onto passengers to complain if they feel uncomfortable and hope someone is listening).
Importantly, the terms of the Copenhagen contracts are such that anyone with a fair grasp of numbers can read and understand them in a matter of hours. This is not least because contracts of this sort are overseen by just one or two people who have other important work to do besides!
The real point of comparing the staffing levels of the ZVV with the Office of the Director of Public Transport, of course, is to show that moving to a Transport Community model will not increase the overall payroll. In fact it will almost certainly cost the Victorian taxpayer less than persisting with the franchising model, and allow some of the millions of dollars of administrative overhead costs to be reallocated to providing actual transport services (including more front-line staff such as tram conductors). Given that many opponents of the government love to draw attention to its level of spending, and spending on public transport in particular, there is an opportunity here to actually save public money and improve service delivery at the same time.
But look at all the responsibilities the Director of Public Transport already has: how is that any different from what the PTUA proposes?
Our position has sometimes been misconstrued as saying the government is not actually allowed to do any tactical planning for public transport. It is then pointed out that under the current franchising model, the Director of Public Transport is still responsible for integrating timetables, planning the network and setting fares, as though this means there is no need to change the governance model. The real problem, however, is that:
- the Directorate is not actually performing these functions very well;
- the Directorate is not performing these functions in a way that serves the public interest; and
- if the Directorate were to undertake tactical planning in a way that reflects best practice and serves the public interest, it would be adopting a highly ‘interventionist’ role which, while taken for granted in other governance models, is contrary to the intent of the franchising model.
The public interest in having a public transport system that provides an effective alternative to car travel in Melbourne is served by such measures as having buses feed into railway stations and coordinate with trains; extending the train network to fill large gaps and serve growth areas; and setting fares at a level that competes with the car. On these and other issues the Directorate has allowed the interests and priorities of the private operators to prevail over the public interest. This is consistent with the role of a public regulator in a franchising model, not with that of a planning agency in a Transport Community model.
The tactical failings of the present system extend right down to issues such as passenger information. A typical example is the inability of Metlink to provide information about bus diversions when road closures occur. When Lonsdale Street had to be closed for several hours in April 2006, passengers seeking information on how to get home were referred to a National Bus information line, which connected to an answering machine.
Another example is the failure of the new electronic displays at Southern Cross Station to provide next-train information on a par with the other City Loop stations. It appears that the displays were provided by the Southern Cross Station Authority according to a specification that is incompatible with the electronic information provided by Connex. Currently, these two service providers are blaming each other for the incompatibility, and the Ombudsman’s office (looking into the matter on request from a PTUA member) considers it outside its scope to force a resolution. As with the electronic SmartBus displays (which failed and had to be reprogrammed at great cost within two years of being installed), it is unclear that the Southern Cross displays will ever be fully operational. Were tactical planning being managed properly by a public authority acting in the public interest, planners would have ensured as a minimum that the specification for the displays was compatible with the train operator’s existing systems.
Legislative changes in 2006 made the Director of Public Transport a ‘referral authority’ for land-use planning decisions, similar to Vicroads’ existing power to intervene in planning schemes. Again, however, there is no indication that these new powers are being used proactively in the interests of greater public transport use.
For example, the most urgent and inexpensive priority for integrating transport and land-use planning is to build new railway stations in growth areas that are currently bypassed by existing train lines (such as Roxburgh Park and Point Cook) and equip them with feeder bus services to penetrate the surrounding residential areas. The PTUA’s 2002 policy document It’s Time To Move identified some 10 locations where new stations are needed—many of these unchanged from a similar document in 1991!—and subsequent urban growth has exposed new inadequacies. Yet in the next 20 years it is proposed to build just four new stations (at Coolaroo, Point Cook, Lynbrook and Pakenham Lakes, the last three not until 2012). The station at Coolaroo, despite costing four times as much as the average new station in Perth, includes no new feeder bus services; only a costly 1200-space car park which will accommodate just 10% of the adult population in the station catchment before filling up. The decision to build this station, like that on Sunday tram services, appears to have been the outcome of lobbying by the local community and the PTUA leading to an election promise from the government, not any proactive planning by the Department.
The other most important priority for transport and land-use integration is the extension of high-frequency bus services into new growth suburbs so that people have access to public transport from the time they move in, before regular car use becomes a habit. While some new bus services have been introduced in fringe suburbs, these run at the same hourly frequencies typical of traditional Melbourne suburban bus services and are clearly intended as ‘charity’ services for non-car-owners rather than as an environmentally friendly alternative for the general population.
In short, the Director’s new status as a referral authority, while removing one of the many disparities between our public transport management organisation and our road management organisation, does not invalidate any of our criticisms of the way the Directorate manages public transport or obviate the need for an independent planning authority.
Even when public transport is provided by a private operator, the nature of the operator’s contract makes a world of difference to the outcomes that matter for passengers. As we have explained, different types of contracts allocate planning responsibilities differently, and so lead to different models of transport governance.
Franchise or ‘net cost’ contracts are an attempt to transfer ‘revenue risk’ to the private operator, by making their remuneration dependent on fare revenue. The franchisee is expected to manage the revenue risk by doing their own tactical planning to some degree. The problem with doing this in a multimodal system is that the operator only has control over their own services, not over the network as a whole, whereas good passenger outcomes depend on a well-functioning network with integrated fares and timetables. When the fare system is multimodal, much of the marginal revenue impact of an operator’s investment in better service isn’t seen by that operator, but is instead distributed to other operators. This means the incentive required for a single operator to improve service can be several times what it would need to be for a central coordinating agency looking to break even across the entire network.
So, while the attempt to transfer revenue risk might succeed in cases where the demand for one operator’s services doesn’t rely on the cooperation of other operators and where the operator has an independent revenue stream, in a multimodal public transport system this approach cannot help but lead to the creation of perverse incentives or disincentives. This is now recognised by planners, particularly in Britain which has experimented the most with this approach. (The more recent franchise contracts in Britain have been made to operate more like gross cost contracts through the use of ‘cap and collar’ arrangements, where the planning authority makes or receives payments to counteract significant differences between actual revenue and forecast revenue.)
The fee-for-service or ‘gross cost’ contracts that are generally used for private operators in the Transport Community model recognise that the system operates as a coordinated entity, by applying revenue and revenue risk to the system as a whole.
Zurich’s public transport succeeds because they’re good at doing timetables, not because of some quirk of bureaucratic organisation.
It’s true Zurich’s planners are among the world’s foremost experts in transport scheduling, but this is far from the only lesson Melbourne can learn from Zurich or from any of the other cities that run successful public transport. Perhaps the most important lesson is the importance these systems place on the passenger, and on the outcomes that matter to passengers. The organisation of institutions under the Transport Community model helps not only with the integration of transport networks but in building immunity against the natural tendency of bureaucracies to turn in on themselves and, in Vuchic’s words, develop the ‘self-defense of incompetence’.
It is also found quite generally that for a system with a long history of decline to make real progress in attracting passengers, a change in management culture is necessary. Even the world’s foremost timetabling experts would be unable to do their job properly within a “can’t-do” institutional culture that resists innovation and operates for the convenience of those who manage the system rather than of those who use it. Cultural change is facilitated by various elements within the Transport Community model, including an independent board, open management, and high public expectations. The fact that the ZVV goes so far as to identify its officers by name on its website says a lot about the culture of Zurich’s public transport.
The ZVV works with eight ‘partner’ operators among the many that provide services, and these ‘partners’ take on some planning tasks such as the production of timetables. This however is a mutual process, where the ZVV determines the principles on which the timetable must be constructed, a timetable is drafted by the operator, and then the timetable is refined through a feedback process. The ZVV is responsible at each stage for ensuring the timetable respects all connections between different services. So the operator’s task is essentially just that of fitting the actual vehicle runs into a predetermined framework.
As the ZVV explain on the English section of their website (ZVV 2007):
Close cooperation with these eight companies makes it possible to ensure that railway, bus, tram and boat workings are optimally coordinated with one another, minimising the time necessary for making connections at scheduled interchanges. The transport-planning section also looks after all infrastructure projects. It makes sure that the various procedures involved in engineering projects run smoothly and that all the statutory and regulatory constraints are complied with.
It may be noted also that in Zurich the train operator, and most if not all of the other ‘partner’ operators, are publicly owned. While there are also private operators, these are smaller and do not take on the ‘partner’ role.
The ZVV runs transport not only in the City of Zurich but in the entire Canton, which includes neighbouring towns, villages and rural hinterland spread over an area not dissimilar in size to Melbourne. Owing to the pattern of high-density towns interspersed with low-density rural areas (all of which are provided with public transport better than that in the average Melbourne suburb) the population of the Canton, while smaller than that of Melbourne, is certainly comparable. Importantly, the Canton’s gross population density of 17.3 persons per hectare is almost exactly the same as that of the Melbourne urban area at 18 per hectare.
Unfortunately the power of PPAs, even when they extend to roads under regional jurisdiction, does not extend to roads that Australian Federal or Canadian Provincial Governments might want to build. So there are a small number of freeways in Vancouver, which are Provincial Highways funded and planned by the Province of British Columbia. Similarly, despite its stance against new freeway construction, the Western Australian Government recently undertook freeway construction work in Perth after being given ‘use it or lose it’ funding by the Federal Government. It is still more politically damaging for a State Government to refuse Federal money than it is to reverse a decision not to build roads.
It is now standard practice for bureaucrats returning from overseas fact-finding tours to report on the overwhelmingly positive view of Melbourne held by overseas planners. The Minister herself is the latest figure to do so, returning from an overseas trip in November 2007 and saying that Melbourne’s system is seen overseas as a very successful one.
Leaving aside the fact that often we only have the Minister’s and her bureaucracy’s word for it that overseas planners think this way, the positive view is easily explained. Viewed from a distance Melbourne’s public transport system really does look impressive. We have more railway track relative to population than just about any other city in the world, the largest tram system outside Europe, rolling stock that is virtually all less than 25 years old and much of it less than 10 years old, and a huge workforce responsible for managing the system. It really is not difficult to paint a very successful picture with the raw statistics.
Even close up, one is also likely to gain a positive impression of the system if approaching it from the perspective of the typical international visitor, who is unlikely to venture more than 5km from the city centre. Short trips focussed on the CBD highlight the advantages of our tram system, at least relative to many other cities where an absence of conspicuous on-street public transport makes travel beyond walking distance difficult.
It is when our public transport is used in the manner of a typical Melbourne resident, to travel from the suburbs to the city or from one suburb to another, that its deficiencies become glaringly apparent. Now and then an international expert experiences this first-hand: as in the case of the visiting professor who took public transport to a Monash University engagement and wound up stranded at Huntingdale station. Or the Toronto planning official who thought to visit a Melbourne colleague at home one Saturday, only to discover when halfway there that the bus they planned on using the rest of the way didn’t run on weekends.
Overseas visitors who experience this side of our system wonder why there hasn’t been a public revolt against the government: they are used to systems where officials lose their jobs for the kind of poor performance we tolerate in Melbourne on a daily basis. More commonly, though, our visitors are driven around town by their colleagues, who know better than to trust the public transport system. As a result very few get to experience public transport in the way Melbourne public transport users experience it from day to day.
So it’s not at all surprising when people with limited experience of our system make positive comments about it. But this has nothing to do with how good our system really is or how it should be run.
Moving to the Transport Community model does not involve recreating the old Met. The chief problem with the old Met was the defensive, inward-looking corporate culture inherited from half a century of failure. The new public agency would be staffed with people who understand how to build patronage and take pride in a well-functioning transport network.
Many of the best public transport cities in the world have moved away from the old nationalised model while retaining public control and in many cases public ownership. Conversely, cities such as Sydney operate under a nationalised model and suffer from the very sources of dysfunction that the Transport Community model is aimed at solving: poor integration of modes, vague division of responsibilities, buck-passing and lack of common branding. We have endeavoured to draw attention to the very important differences between a nationalised model and a Transport Community model so that our preferred option can be distinguished from historical arrangements that the PTUA has soundly criticised in the past.
Figures for cancellations and late running were generally lower in the first few years of privatisation than in 1998/99 immediately before the rail and tram systems were privatised. However, the years 1998 and 1999 were notorious for poor performance, due to operational difficulties associated with the split of the system into five franchise areas. If the figures after privatisation are compared instead with the figures for 1996/97 before the institutional tinkering began, the story is rather different.
For example, in 2001/02 the rate of tram cancellations was 0.4%. This is an improvement on the rate in 1998/99 which was 1.1%; however, it is poorer than the rate in 1996/97 of 0.2%. The rate of train cancellations was 0.5% in 2001/02, improving upon the rate of 1.0% in 1998/99 but not on the rate of 1996/97 which was also 0.5%. Similar trends are apparent in the figures for late running, but are complicated by the fact that the definition of late running has changed since 1997.
Here too, we see the effect of ‘regulatory capture’ of the government by the operators, with a new timetable in 2007 that ‘solves’ the late running problem for trains by allowing them to run slower!
The ‘privatisation debate’ is a distraction that is holding us back from pursuing real improvements to public transport.
On the contrary, it is quite clear that for many years Victoria’s public transport has been ‘bogged down’ by an institutional inertia that is by now well understood and documented in widely-used textbooks like Vuchic’s Urban Transit. The 340 employees of the Office of the Director of Public Transport, if they were of a mind to do so, could of course carry out important improvements to public transport at the same time as debating a few community activists. Elsewhere, such debates are an integral part of both strategic and tactical planning processes, and are regarded as an important avenue for democratic oversight of planning functions.
The argument of this paper has been that real improvements to public transport are unlikely to occur without reform of its governance arrangements. The current arrangements have been in place for over seven years, and in that time the only significant patronage increases have been due to fortuitous external factors (the Commonwealth Games, high petrol prices and growth in population and employment). Both the private operator and the regulator responsible for defending the public interest are hostile to the idea of even modest rail extensions being undertaken within the next 15 years, and instead are directing huge sums of money toward ‘capacity’ problems, such as that of the Dandenong line, that have relatively inexpensive tactical solutions (which often can be found simply by examining the way the same system was run 20 or 30 years ago). Even under the $7 billion of expenditure promised in the Meeting Our Transport Challenges strategy, most people are unlikely to see any reason (other than high petrol prices) to swap the car for public transport before 2020. And the most basic sign of a coordinated network—buses timetabled to meet trains at stations—is still almost entirely absent, with only vague plans to redress the situation. There can be little doubt that some kind of change is needed.