Annual taxi licence release for Sydney 2013/14 › files › sharedassets › ... · Annual taxi licence release for Sydney 2013/14 . Contents . 1 Introduction : 1 ... the likely

Independent Pricing and Regulatory Tribunal

Annual taxi licence release for Sydney 2013/14

Transport — Final ReportFebruary 2013

Annual taxi licence release for Sydney 2013/14

Transport — Final Report February 2013

ii IPART Annual taxi licence release for Sydney 2013/14

© Independent Pricing and Regulatory Tribunal of New South Wales 2013 This work is copyright. The Copyright Act 1968 permits fair dealing for study, research, news reporting, criticism and review. Selected passages, tables or diagrams may be reproduced for such purposes provided acknowledgement of the source is included.

ISBN 978-1-922127-65.5

The Tribunal members for this review are: Dr Peter J Boxall AO, Chairman Mr James Cox PSM, Chief Executive Officer and Full Time Member Mr Simon Draper, Part Time Member

Inquiries regarding this document should be directed to a staff member: Jennifer Vincent (02) 9290 8418 Ineke Ogilvy (02) 9290 8473

Independent Pricing and Regulatory Tribunal of New South Wales PO Box Q290, QVB Post Office NSW 1230 Level 8, 1 Market Street, Sydney NSW 2000 T (02) 9290 8400 F (02) 9290 2061 www.ipart.nsw.gov.au

http://www.ipart.nsw.gov.au/

Contents

iii IPART Annual taxi licence release for Sydney 2013/14

Contents

1 Introduction 1 1.1 Purpose of this review 2 1.2 Our approach to recommending the number of new licences 3 1.3 Estimated outcomes of the recommended number of licences 4 1.4 Final recommendations in this report 6 1.5 How the review was conducted 7 1.6 Differences between draft and final report 7 1.7 Structure of this report 8

2 Overview of the Sydney taxi industry 9 2.1 Taxi industry structure 9 2.2 Types of licence in the Sydney taxi market 13 2.3 Number of licences by type in Sydney taxi market 15

3 Context for the review: objectives and factors 17 3.1 Circumstances that led to the 2009 amendments 17 3.2 Factors in the 2009 amendments and other objectives we must consider 18 3.3 Approach used to determine number of new annual licences to be

released to date 22 3.4 Success of this approach in meeting the amendment’s objectives to date 23

4 The demand for taxi services 28 4.1 How changes in demand relate to the change in licence numbers 29 4.2 Estimating changes in demand due to external factors 29 4.3 Estimating latent demand 34 4.4 Measuring the overall change in demand based on taxi trip data 39

5 The number and mix of licences to be released 42 5.1 Balancing a more affordable means of entry into the taxi market with

avoiding unreasonable impacts on existing licence-holders 44 5.2 Determining the mix of unrestricted and Peak Availability Licences

(PALs) 48 5.3 Estimated effects of releasing 160 unrestricted licences and 200 PALs in

2013/14 54 5.4 Could the outcomes be different from our estimates? 65

Contents

iv IPART Annual taxi licence release for Sydney 2013/14

6 Making an allowance for replacement licences and Wheelchair Accessible Taxi licences 69 6.1 Additional licences that need to be released 69 6.2 What should be done to help operators relinquish a more expensive

licence in favour of a cheaper one 71

7 Process for setting a price and restrictions on who may hold licences 74 7.1 Tender process 74 7.2 The need for a reserve price 76 7.3 Reserving some licences for drivers and/or restricting the number that

can be granted to the same applicant 77

Appendices 81 A Terms of Reference 83 B List of submissions 85 C Taxi industry model 88

Glossary 91

1 Introduction

Annual taxi licence release for Sydney 2013/14 IPART 1

1 Introduction

By 31 March each year, Transport for NSW must determine how many new annual taxi licences it will release in Sydney from 1 July that year.1 To assist Transport for NSW to make its next determination, the NSW Government has asked the Independent Pricing and Regulatory Tribunal of NSW (IPART) to review and make recommendations on the required number of new licences for the year commencing 1 July 2013.

Our recommendation is that 225 unrestricted annual licences, 230 Peak Availability Licences (PALs) and 1 Fringe Area (Richmond/Windsor) Licence should be released, comprising:

160 new unrestricted licences and 200 new PALs to allow for growth and peak demand

an additional 70 unrestricted licences and 30 PALs to replace licences we expect to be handed back or not renewed (including new licences effectively swapped for existing licences to get a better price)

an adjustment to subtract 5 licences to take into account the expected issue of additional Wheelchair Accessible Taxi licences, outside the annual licence tender process

1 Fringe Area (Richmond/Windsor) Licence to replace one that was not renewed in September 2012.

We are recommending a slightly different balance of unrestricted licences and PALs compared to our draft report.

1 Excluding licences for Wheelchair Accessible Taxis, which are available on demand. See Section

32C of the Passenger Transport Act 1990.

1 Introduction

2 IPART Annual taxi licence release for Sydney 2013/14

1.1 Purpose of this review

Prior to 2009, the number of taxi licences in Sydney did not keep pace with the growth in Sydney’s population, household and business income, economic activity or tourism numbers. This meant that there were not enough taxis on the road to meet passenger demand. As a result, taxi licences became a scarce commodity and were very expensive to buy and lease. This increased the costs of operating a taxi business as well as fares for taxi services. In 2008/09 taxi licences were around $365,0002 to buy outright, or around $26,000 (not including GST)3 to lease for a year, making up around 20% of the costs of providing taxi services and fares paid by passengers.

To address these problems, the Passenger Transport Act 1990 (the Act) was amended in 2009. The amendments included the introduction of Transport for NSW’s determinations of the number of new Sydney annual taxi licences to be released each year. Between the end of 2009 and the end of 2012, the Sydney taxi fleet has increased in size by around 8%.4 However, lease costs for licences leased through networks continued to increase in real terms (that is, by more than the rate of inflation) to the end of 2011, followed by a slight drop during 2012.5 It is not significantly more affordable to enter the taxi market than it was at the time of the reforms in 2009.

IPART has been asked to review and recommend the number of new taxi licences that should be released in 2013/14. We must consider the objectives of the 2009 amendments, which were to:

ensure the supply of taxis responds closely to growth in passenger demand balance the need for more affordable entry into the taxi market with the need

to avoid unreasonable impacts on existing licence holders

reduce barriers to entry and encourage competition place downward pressure on fares over time, and simplify existing taxi licence structures.

2 Average licence transfer value for 2008/09 (in 2008/09 dollars) based on information from

Transport for NSW. 3 In 2008/09 dollars. IPART calculation, based on the CIE survey data 2011 deflated by NSW

Taxi Council information about lease prices from a sample of taxi networks provided for IPART fare reviews between 2008 and 2012.

4 Information provided by Transport for NSW for 1 January 2010 compared to 1 January 2013. 5 Information about lease prices from a sample of taxi networks received from the NSW Taxi

Council on a commercial-in-confidence basis.

1 Introduction


We must also consider the factors in section 32C(3) of the Act, which are:

the likely passenger demand for taxi-cab services, including latent demand the performance of existing taxi-cab services the demand for new taxi-cab licences

the viability and sustainability of the taxi-cab industry, and any other matters considered relevant, having regard to the objective of

ensuring improved taxi-cab services.

(The full terms of reference for our review are provided at Appendix A.)

1.2 Our approach to recommending the number of new licences

We firstly considered the external sources of demand for taxi services, such as population, tourism, household and business income and business activity. The range of indicators suggests that demand from these external factors is likely to grow by 2.5% per year on average. We recommend for this review and the next 4 annual reviews of taxi licences that the minimum increase in the number of licences is 2.5%, reflecting a longer term estimate of growth in demand for taxi services resulting from external factors. For simplicity and certainty we consider that the 2.5% should be applied to the current number of taxi licences of 5,647,6 so that a minimum of 140 new unrestricted licences is released each year.

However, if the growth in licence numbers simply keeps pace with the growth in demand caused by external factors, lease costs are likely to be maintained in real terms.7 This would mean that many of the objectives of the amendments would not be met – it will not become more affordable to provide taxi services, so no downward pressure would be placed on fares, and there would be no improvements in the waiting times for customers.

On the other hand, if the growth in licence numbers greatly exceeds the growth in demand from external factors, the income for existing licence owners from the lease payments they receive would fall substantially. Our terms of reference require us to balance the 2 objectives of providing more affordable entry to the market for operators and avoiding unreasonable impacts on licence holders.

Determining how many additional licences to release over and above the base level of demand cannot be accomplished by a simple formula. The taxi industry consists of complex interactions between supply of taxis, demand for taxi services, waiting time and fares. Different classes of licence (such as PALs) also have a different impact on the way the market works.

6 As at 1 January 2013. 7 Assuming fares and all other costs of operating taxi services increase in line with inflation.

1 Introduction


1.3 Estimated outcomes of the recommended number of licences

We used a model of the taxi industry developed by the Centre for International Economics (the CIE) to help us to understand the likely outcomes of adding different numbers and types of licence to the Sydney taxi market.8 The model estimates the outcomes in terms of changes to passenger demand for taxis, percentage of time a taxi spends occupied, waiting time for passengers and likely licence lease costs.

The model is a long-term equilibrium model. That is, it shows what the estimated outcomes are when the market has fully adjusted to the changes brought about by additional licences. It does not map out the transition path, or the time it takes to achieve the outcomes, but it is clear about the direction of the path. In short, our recommended increase in the number of licences will ensure that the supply of taxis responds closely to the growth in passenger demand, reduce barriers to entry into the taxi market, and improve services to customers through reduced waiting times.

We consider that releasing 225 unrestricted annual licences and 230 PALs is the combination that best meets the requirement to consider demand and latent demand and provides an appropriate balance between: positive outcomes for passengers (reduced waiting times and more taxi trips),

driver income (taxi occupancy and driver hourly earnings), operator costs (lower annual licence costs)

impacts on existing licence-holders (lower income from leasing out licences).

We also consider that the mix of licences we are recommending will help to mitigate transition effects.

1.3.1 A mix of PALs and unrestricted licences improves benefits for passengers and drivers

Releasing a mix of PALs and unrestricted licences produces larger passenger benefits for a given impact on licence owners: peak waiting times are reduced significantly compared with the release of only unrestricted licences.

PALs also provide a more affordable opportunity for drivers to invest in their own taxi business: without low cost PALs it is not viable for drivers to drive their own cab without also hiring other drivers (and it is likely to be that way for some time).

The number of licences we have recommended will also increase demand for driver labour, which should have a positive effect on driver incomes.

8 See Appendix C for information on the Taxi Industry Model. The model is also available on our

website, www.ipart.nsw.gov.au.


1 Introduction


1.3.2 The impact on licence owners is not unreasonable

We expect our recommendations to reduce income from perpetual licence leases by around 10%, all other things being equal.9 Some reduction is necessary in order to achieve the objective of more affordable entry into the taxi market. We consider that an initial reduction in licence lease costs of 10% provides an appropriate balance between improved affordability for operators and reduced income for licence owners. Smaller impacts do not provide much in the way of benefits to customers; the reduction in waiting times under our recommendations is twice what it would be under an estimated 5% reduction in lease income.

This is in the context of a reduction of licence lease costs of 25% over the next 5 years. In the event that we are given a referral to recommend licence numbers for 2014, we would re-evaluate the context, taking account of the experience of the first year.

1.3.3 What will be the transition effects on the industry?

The transition process largely depends on the timing of the adjustment of lease costs. The NSW Taxi Council has submitted that lease costs will be slow to adjust and drivers and operators will go out of business before lease costs adjust. However, it is quite feasible that lease costs will adjust more quickly than envisaged by the NSW Taxi Council, and we consider that the Taxi Council could assist the industry to make a quicker transition.

Issuing more licences as PALs should also help to accelerate the transition to lower cost licences, by offering an alternative, more affordable business model to operators immediately. It will bring forward greater benefits for passengers (in the form of shorter waiting times at peak periods) without additional impacts on existing licence owners.

There will be the opportunity to review progress when the determination of annual taxi licences for 2014/15 commences. However, given that new licences for the current determination will only commence from the second half of 2013, only a preliminary assessment of any changes to the industry will be possible.

Nevertheless, it is important to emphasise that transition effects are by definition temporary and that the change is in the long term interests of the industry as a whole.

9 That is, fares stay at their current levels, while all other costs of operating taxi services increase

at the same rate as the cost of living.

1 Introduction


1.4 Final recommendations in this report

Transport for NSW should release 225 unrestricted annual licences, 230 Peak Availability Licences (PALs) and 1 Fringe Area licence for tender in 2013/14, comprising:

160 new unrestricted annual licences and 200 new peak availability licences (PALs) 43

65 unrestricted licences and 30 PALs to replace licences we expect to be handed back or not renewed. This includes an adjustment to account for the expected issue of additional Wheelchair Accessible Taxi licences outside the annual tender process. 70

1 Fringe Area (Richmond/Windsor) Licence 70

Transport for NSW should retain the current tender process:

Transport for NSW should continue to use the existing tender process, that is, a sealed electronic tender, with pay-as-bid prices for successful tenderers. 75

Transport for NSW should continue to release licences in accordance with the same timetable followed in previous years. 76

Transport for NSW should publish the provisional results of tenders for annual licences (for example, by publishing the median and mean successful bids) within 2 weeks after tenders have closed. 76

No reserve price should be set for licences tendered. 77

No restrictions should be placed on who may bid for licences or on the number of licences that any one person may hold. 79

Other recommendations:

Transport for NSW should investigate the cost and feasibility of mandating a regulator data set. 40

Roads and Maritime Services should require authorised Sydney taxi networks to provide reports on utilisation of all affiliated taxis for a sample of 4 weeks per year commencing in June 2013. 41

Transport for NSW should ensure that its administration processes and tender documentation support annual licence churn as a positive mechanism operators can use to reduce their annual licence costs. 73

Transport for NSW should remove any unnecessary administrative costs of churn, such as the requirement for a blue slip and deregistration/registration of the same vehicle with a new licence. 73

1 Introduction


1.5 How the review was conducted

We released an issues paper in October 2012, which set out the key issues we had identified as part of the review, and sought comment from interested parties.

Submissions were due by 9 November 2012. We received 42 submissions. Prior to the due date for submissions we held a public roundtable on 24 October to provide stakeholders with a further opportunity for input.

We released our draft report on 10 December 2012, with submissions due by 21 January 2013. We received 66 submissions.

Appendix B contains a list of all submissions received.

All the publications associated with the review, including reports, submissions and a transcript of the public roundtable, are available on our website, www.ipart.nsw.gov.au. A copy of the taxi industry model developed for the review is also available on our website.

1.6 Differences between draft and final report

In our draft report, we recommended slightly more PALs be released (280 compared to 230) and slightly fewer unrestricted licences (205 compared to 225).

We have also added 2 recommendations following consideration of submissions and discussion with Transport for NSW: That Transport for NSW remove any unnecessary administrative costs of

churn, such as the requirement for a blue slip and deregistration/registration of the same vehicle with a new licence.

That Transport for NSW publish the provisional results of tenders for annual licences (for example, by publishing the median or mean successful bids) within 2 weeks after tenders have closed.


1 Introduction


1.7 Structure of this report

This report sets out our decisions and recommendations. It is structured as follows: Chapter 2 provides an overview of the taxi industry in Sydney

Chapter 3 provides the context for the review, including the objectives and factors in our terms of reference

Chapter 4 explains how we have forecast the change in demand for taxi services

Chapter 5 explains our recommendation on the number and mix of new licences to be released for 2013/14 and presents the likely impacts on the various segments of the taxi industry

Chapter 6 considers what adjustment needs to be made to allow for some licences that will be handed back to Transport for NSW (or not renewed) and to allow for the Wheelchair Accessible Taxi licences that may also be issued by Transport for NSW

Chapter 7 explains our decisions on whether new licences should be released by auction or sealed tender, and how the price of those licences should be set, as well as whether some licences should be reserved for drivers and whether restrictions should be placed on how many can be issued to one applicant.

2 Overview of the Sydney taxi industry



Sydney’s taxi industry involves many participants, as well as a range of different taxi licence types. To help stakeholders understand the context for this review, the sections below provide an overview of the industry’s structure, and the licence types and numbers available.

2.1 Taxi industry structure

The taxi industry involves a range of participants with different roles and responsibilities – including regulators, networks, operators, drivers and licence owners. Some of these participants play several roles within the ‘supply chain’ for taxi services. For example, many operators are also drivers, networks may own licences, and some licence-owners also operate and drive a taxi.

Figure 2.1 below provides an overview of the taxi industry structure in NSW. The following sections outline the key participants’ roles and responsibilities.

2.1.1 Regulators

While taxi-cabs are a privately (rather than publicly) owned mode of transport, taxi services are usually regulated by government to ensure they meet acceptable safety and quality standards. In NSW, the main regulator is Transport for NSW.

Transport for NSW’s role includes issuing taxi driver authorities, taxi operator accreditations and taxi licences. It also authorises taxi networks, and enforces taxi service standards for networks, vehicles, operators and drivers. In addition, Transport for NSW is responsible for setting taxi fares.

IPART reviews and recommends changes in taxi fares each year. However, Transport for NSW makes the final fare decisions. For the first time, we have also been asked to review and recommend the number of new taxi licences to be released. Again, Transport for NSW will make the final decision.



Figure 2.1 Overview of the taxi industry in NSW

Source: The CIE, Reweighting of the taxi cost index – final report, April 2012, p 14.

2.1.2 Taxi networks

All taxis must be affiliated with a taxi network.10 Taxi networks provide a radio booking service to the taxi operators who are affiliated with them, as well as security monitoring services for taxi drivers and passengers. In some cases, they also provide additional services to operators and drivers, such as training, leasing or sub-leasing taxi licences, insurance broking, and repairs and maintenance. In addition, they monitor and enforce service standards for operators, drivers and vehicles. Networks must be authorised by Transport for NSW.

There are currently 12 networks operating in Sydney, some of which are linked to each other through their business structures.11

10 Passenger Transport Act 1990, S 30(1). 11 Information provided by Transport for NSW.

PROVIDERS OF TAXI SERVICES

Licence plate holders

Operators Drivers

Passengers

Phone and internet bookings

Annual network fees

Licence plate lease payments

Pay-ins or revenue sharing

Booking dispatch

Fares

Taxi services

Transport for NSW

Sets network

standards

Monitors/ enforces standards

May issue new licences

Payment for new licences

Networks

Sets and enforces

regulations



2.1.3 Taxi operators

Taxi operators are responsible for the day-to-day management of one or more taxi cabs. Operators may be individuals or corporations, and must be accredited by Transport for NSW. They must also be affiliated with an authorised network, and are required to fit out their vehicle with that network’s livery and install the network’s communications equipment. They must insure and maintain their vehicle(s).

In addition, operators must hold a taxi licence for each vehicle they operate. Licensing entry to an industry is a common method of ensuring that participants are qualified to be in the industry, and that they maintain standards of safety and quality. However, the taxi industry is one of the few industries where the number of operators is restricted by the number of licences issued.

The cost to an operator of holding a taxi licence varies, depending on whether:

the operator owns or leases the licence the type of licence it is the source of the licence (secondary market or Transport for NSW), and

the market conditions at the time of obtaining the licence (and any change to market conditions subsequently, for leased licences).

Currently, some 60% of operators lease their taxi licences, while the other 40% either own them or hold the new annual licences issued by Transport for NSW since 2010.12 As at 1 January 2013, there were 2,609 active accredited operators in Sydney, many of whom operate just one taxi.13

Many operators are individuals who also drive their own taxis. In Sydney, most operators also arrange for other drivers to drive their taxis by charging the driver a fixed fee to take out (“bail”) the taxi.14 The maximum fee that the operator can charge the driver is set per shift by the Industrial Relations Commission, and currently ranges from $266.55 for a Friday or Saturday night shift, down to $175 for all day shifts.15

12 The CIE, Reweighting of the taxi cost index – final report, April 2012, p 44. 13 Information provided by Transport for NSW. 14 This arrangement is determined by the Industrial Relations Commission through the Taxi

Industry (Contract Drivers) Contract Determination 1984. The fixed fee arrangement is known as “Method 2.” Alternatively, operator and driver can agree to divide the taxi fares collected on the shift – usually 50/50 (“Method 1”). While a shared-fare arrangement is common in other states in Australia, it is rarely used in Sydney.

15 See www.industrialrelations.nsw.gov.au/biz_res/oirwww/pdfs/Awards/Award_0103.pdf.

http://www.industrialrelations.nsw.gov.au/biz_res/oirwww/pdfs/Awards/Award_0103.pdf



2.1.4 Taxi drivers

A taxi driver must be licensed to drive in NSW and authorised by Transport for NSW. Drivers must wear the approved uniform of the network to which their vehicle is affiliated, and be logged in to that network while available for hire. Drivers are responsible for paying for fuel and car washing under bailment Method 2 (fixed pay-in).

As at 1 March 2012, there were 18,791 authorised drivers in Sydney, although not all were active.16

2.1.5 Licence owners

Although a taxi operator must hold a licence (either by owning or leasing it) in order to conduct a taxi business, there is no requirement for a taxi licence owner in NSW to play any part in providing taxi services – and many do not. This means that taxi licences are comparable to other financial assets. The expected profits of owning a licence are determined by how profitable it is to either operate a taxi now and into the future, or how profitable it is to lease the licence to another party.

There are currently 5,647 licences in the Sydney taxi market (as at 1 January 2013).17 Most of these licences are owned by individuals, and more than half are owned by individuals with only 1 licence each. The taxi networks own 446 licences between them.18 The taxi networks also manage licences for individual owners by organising to lease or sub-lease them to operators.

Transport for NSW advises that about 75% of taxi licences are operated by someone other than the licence owner.19 This includes licences owned by taxi networks and operated by affiliated operators. The NSW Taxi Council pointed out in its submission on the draft report that this also includes operators who ‘are registered under a different name to the entity registered as their licence owner (eg, their own superannuation fund, family trust or company) despite the fact that the operator is the beneficial owner of the licence’.20 Nevertheless, we remain of the view that a large proportion of owners hold a licence simply as an investment, whether or not they were previously involved in the taxi industry (such as retired taxi operators).

16 Information provided by Transport for NSW for the 2012 taxi fare review. 17 Information provided by Transport for NSW. Transport for NSW has advised that 55 of the

licences are ‘on hold’ – that is, the licence is in force, but a taxi is not currently being operated using the licence.

18 Information provided by Transport for NSW. 19 Transport for NSW information return for 2012 review of fares, ‘Leased taxis as at 1 March

2012’. 20 NSW Taxi Council submission, 21 January 2013, p 29.



2.2 Types of licence in the Sydney taxi market

A range of different types of taxi licence are currently valid in NSW. These include perpetual licences, ordinary licences, short-term licences and annual licences. In addition, some licences are unrestricted, while others have restrictions on when, where and how the taxi can be operated.

2.2.1 Perpetual licences

Prior to 1990, perpetual taxi licences were issued to drivers from time to time by ballot. At first, the transfer (ie, sale) of these licences was permitted only to another driver and under certain conditions (eg, only after 10 years or in case of ill-health or death). However, the restrictions on their transfer were gradually removed, and in the 1980s perpetual licences became fully tradeable.

New perpetual licences have not been issued in Sydney since 1990. However, existing licences can still be obtained on the secondary market. In 1989, the average transfer value for an unrestricted perpetual licence was $164,000.21

2.2.2 Ordinary licences

From 1990, ordinary licences with 50-year terms were issued by the transport department on request at the prevailing price for perpetual licences. These licences came with the right to renew them at the end of their term, so they are effectively perpetual. However, the market assessed ordinary licences as less valuable than perpetual licences (possibly because of the prospect of a renewal fee being charged). As a result, prospective licence owners preferred to purchase perpetual licences in the secondary market.

New ordinary licences have not been issued in Sydney since 2009.22 However, existing licences can still be transferred on the secondary market. Existing licences now transfer at prices similar to perpetual licences. During 2012 the average transfer price for ordinary and perpetual licences was $410,000.23

21 Information provided by Transport for NSW. 22 But they continue to be available from Transport for NSW on application for areas outside

Sydney. 23 Information provided by Transport for NSW.



2.2.3 Short-term licences

From the late 1990s, some short-term licences with a term of up to 6 years were issued. Their owners can lease them to another person, but cannot transfer them, and they expire at the end of their term. New short-term licences have not been issued in Sydney since 2009, and as they expire they are being replaced by new annual licences.24 No short-term licences will expire during 2013/14.

2.2.4 Annual licences

In December 2009, taxi licensing legislation was amended to introduce an annual review and determination of the number of new licences (other than Wheelchair Accessible Taxi (WAT) licences) to be released in the Sydney Metropolitan Transport District by tender or auction each year. The licences are annual, and are automatically renewable for a term of up to 10 years. The owners of these licences can lease but not transfer them.

Transport for NSW can issue the licences to increase the number of licences in the market as well as to replace short-term licences as they expire. Transport for NSW must formally determine by 31 March each year how many new annual licences to release from 1 July that year. This year IPART is reviewing and recommending to Transport for NSW how many new licences should be released from 1 July 2013.

2.2.5 Unrestricted and restricted licences

While some taxi licences are issued without restrictions, others have restrictions on them. The most common types of restricted licences are: Peak Availability Licences (PALs) (have restricted hours of operation).

Fringe Area Licences (must be operated predominantly in nominated geographical areas on the outskirts of Sydney).

Wheelchair Accessible Taxi (WAT) licences (must be operated to give preference to transporting wheelchair users).

PALs can only be operated between 12 noon and 5 am, and are typically driven for a single shift per day. This means that many of these taxis are on the road at 3 pm and 3 am changeover times, which helps address the shortage of taxis on the road at these times.

24 But they continue to be available from Transport for NSW on application for areas outside

Sydney.



Taxis with a Fringe Area Licence are authorised for hire only within the area of operation. They can only accept hires that originate outside their area of operation where the hiring has been pre-booked and the person is being taken to a place within the area of operation.

For the Sydney Metropolitan Transport District, new and replacement PALs and Fringe Area Licences are now issued through the annual release process, and are part of IPART’s review. Annual WAT licences will continue to be available from Transport for NSW on application for $1,000 a year (and therefore do not form part of our review).

2.3 Number of licences by type in Sydney taxi market

Currently there are 5,647 taxi licences in the Sydney market as at 1 January 2013 compared to 5,231 as at 1 January 2010. The taxi fleet has increased by 8%, comprised of:

net 310 additional unrestricted licences since January 2010, increasing the taxi fleet by 5.9%

net 21 additional PALs, increasing the taxi fleet by 0.4%

net 90 additional WAT licences, increasing the taxi fleet by 1.7%.

Figure 2.2 shows the change in licence numbers and composition in Sydney since November 2008.

Figure 2.2 Change in licence numbers and composition, Sydney taxi market

Data source: Transport for NSW.



As shown above, most licences in operation are unrestricted. PALs currently make up around 5% of all licences in the Sydney market. Fringe Area Licences make up a smaller part of the total fleet, with 14 Fringe Area Licences currently being operated.25

25 A Fringe Area Licence was handed back in September 2012 and we propose that a replacement

should be tendered in 2013/14; 2 WAT licences were reclassified as Fringe WATs in December 2012. (Information provided by Transport for NSW).

3 Context for the review: objectives and factors



As previous chapters indicated, the current approach to releasing new taxi licences in Sydney has been in use since December 2009, when the Passenger Transport Act 1990 (the Act) was amended.26 The amendments establish that each year, Transport for NSW must review and determine the number of new licences (other than Wheelchair Accessible Taxi (WAT) licences) to be released in the Sydney Metropolitan Transport District by tender or auction each year.

To help stakeholders understand this approach, and what it – and our review – is intended to achieve, the sections below discuss: the circumstances that led to the 2009 amendments to the Act

the factors that we must consider in recommending the number of new licences to be released in 2013

the approach used to determine this number in 2010, 2011 and 2012

the success of this approach in meeting the amendment’s objectives to date.

3.1 Circumstances that led to the 2009 amendments

From around 1980 to 2009, growth in the number of taxi licences in Sydney did not keep pace with the growth in the city’s population, household and business income, or economic and tourism activity. The restriction on the supply of taxis meant that licences became a scarce commodity, which pushed up the costs of buying or leasing a licence.

In general, this situation benefited the owners of these licences, as it increased the return they could earn on their investment. However, it imposed additional costs on other participants in the taxi industry, and ultimately on passengers. For example, by 2008/9 operators were required to pay around $26,000 per year (not including GST) in lease costs just to enter the market27 (or invest $365,000 if they chose to purchase a licence).28 Then they had to collect enough revenue through

26 Amendments were made by the Passenger Transport Amendment (Taxi Licensing) Act 2009,

No 118. 27 In 2008/09 dollars. IPART calculation, based on the CIE survey data 2011 deflated by NSW

Taxi Council information about lease prices from a sample of taxi networks provided for IPART fare reviews between 2008 and 2012.

28 Information from Transport for NSW for 2008/09 transfer values (in 2008/09 dollars).



fares or pay-ins from drivers to cover that cost, so pay-ins were higher than they would otherwise have been. In addition, fares had to be set high enough to provide sufficient revenue for drivers to afford the pay-ins. Passengers paid the cost of these higher fares. With the cost of a licence lease amounting to around 20% of the cost of keeping a taxi on the road for a year,29 potentially $4 of every $20 fare went to pay for the licence lease. Passengers also paid the cost of the imbalance between the supply and demand for taxis – for example, by experiencing longer waiting times for taxis, or not being able to get a taxi when they wanted one.

3.2 Factors in the 2009 amendments and other objectives we must consider

The intention of the 2009 amendments was to reduce the effects of past restrictions on the supply of taxi licences described above. This is clearly reflected in our terms of reference, which list the objectives of the amendments and the factors in the legislation that we must consider in making our recommendation.

3.2.1 Objectives of the amendments

The terms of reference for this review list the objectives of the 2009 amendments as:

ensuring the supply of taxis responds closely to growth in passenger demand balancing the need for more affordable entry into the taxi market with the

need to avoid unreasonable impacts on existing licence holders

reducing barriers to entry and encouraging competition placing downward pressure on fares over time simplifying existing taxi licence structures.

Ensuring the supply of taxis responds closely to the growth in passenger demand

This objective suggests that the number of new licences released should ensure that the supply of taxis increases as the demand for taxi services grows.

This suggests the approach used to determine this number must include estimating how demand for taxi services changes over time. Passenger demand for taxi services is driven by factors that include population, household and business income, economic activity, and tourism numbers. Growth in these is likely to increase the size of the potential market of taxi users, and increase the frequency with which people use taxi services.

29 The CIE, Reweighting of the taxi cost index – final report, April 2012, p 9.



In addition, the factors that must be considered in determining the number of new licences to be released specify that the consideration of passenger demand must include not only likely (or actual) demand, but also latent demand (see section 3.2.2 below).

Balancing the need for a more affordable means of entry into the taxi market with the need to avoid unreasonable impacts on existing licence holders

The findings of the CIE’s 2011 survey of drivers and operators conducted for IPART’s taxi fare review indicate that the average cost of leasing a taxi licence for a standard urban taxi in Sydney was $28,789 (excluding GST) per year at the time of the survey (4th quarter of 2011), or around 20% of the total cost of keeping a taxi on the road for a year.30 The other costs include driver labour, fuel, insurance, car lease costs, maintenance, operator labour and network fees.

To update the CIE’s estimated lease cost from 2011 to the current year, we have compared the average annual licence fee for all licences on foot as at 31 December 2012 with the annual average licence fee for all licences on foot as at 31 December 2011.31 We consider that this is likely to be a reliable indicator of the general movement in market lease rates. The change from 2011 to 2012 was -2.2%, which gives an estimated average lease cost as at the end of 2012 of $28,155.

As licence leases derive their value from their scarcity, increasing the supply of licences should reduce lease costs. However, if the growth in licence numbers simply keeps pace with the growth in demand from external factors, lease costs are likely to be maintained in real terms.32 In order to achieve the objective of a more affordable means of entry into the taxi market, more licences need to be released than the number that is required to meet the annual increase in passenger demand.

On the other hand, if the additional licence numbers greatly exceed the growth in demand, lease costs will fall substantially, which means a bigger impact on the incomes of existing owners of perpetual licences. Our terms of reference require us to balance the two objectives of providing more affordable leases for operators and avoiding unreasonable impacts on licence holders.

Reducing barriers to entry and encouraging competition

We consider the main barrier to entry is the cost of taxi licences, discussed above. However, other structural issues might also affect entry into the taxi industry and competition within it. These include the number of licences one person can hold; how a person obtains a licence (holding it themselves, or leasing it from an owner 30 The CIE, Reweighting the Taxi Cost Index Final Report, March 2012, p 9. 31 Information provided by Transport for NSW. 32 That is, assuming fares and all other costs of providing taxi services increase in line with

inflation, lease costs would also increase in line with inflation.



or through a third party); and whether licence owners are involved in the industry as drivers, operators, licence owners or as 2 or 3 of these roles.

Placing downward pressure on fares over time

As Chapter 2 noted, we review and recommend taxi fares to Transport for NSW each year. In recent years, we have based our recommended fare increases on the annual change in the cost of providing taxi services as measured by a Taxi Cost Index. As part of our 2012 review of taxi fares we decided to set the licence lease cost inflator to zero. As part of our next 2013 fare review we will reconsider this, as increasing the supply of taxi licences and thus reducing lease costs should put downward pressure on fares.

Simplifying existing taxi licence structures

Existing short-term licences are gradually being replaced as they expire by new annual licences. While no new ordinary or perpetual licences will be issued in Sydney, existing ones will not be replaced or revoked, so there is a limit to the extent to which the release of new licences can simplify existing taxi licence structures.

3.2.2 Factors to be considered in determining the number of new taxi licences per year

Section 32C(3) of the Act indicates that in determining the number of new annual licences to be released in any year, the following factors must be considered: the likely passenger demand for taxi-cab services, including latent demand

the performance of existing taxi-cab services the demand for new taxi-cab licences the viability and sustainability of the taxi-cab industry

and any other matters considered relevant, having regard to the objective of ensuring improved taxi-cab services.

Likely passenger demand and latent demand for taxi-cab services

As noted above, demand for taxis is driven by a range of factors, including population, household and business income, economic activity, and tourism numbers. We could also directly observe demand for taxis by looking at the number of taxi trips taken in a year.



But as most people see taxis as a discretionary service, there is likely to also be latent demand for taxi travel – that is, demand that we cannot directly observe. This includes the demand by people who would have liked to travel by taxi but didn’t. For example, they might have thought taxi travel was too expensive or the waiting time would be too long or the taxi might not turn up, and so made alternative arrangements, such as driving their own car, catching public transport, or booking a hire car instead. Alternatively, they might have decided not to travel at all.

We commissioned Taverner Research to conduct a survey of 2000 Sydney residents to help us understand the behaviour of taxi users to inform our estimates of responsiveness to price and the time taken to catch a taxi.33 The results of the survey were published at the same time as our taxi licence review draft report.

The performance of existing taxi-cab services

Performance of existing taxi services could be considered an indicator of the balance between supply and demand. For example, increases in the numbers of trips, length of waiting time, or number of bookings where there is ‘no car available’ to complete the job could indicate that demand has grown (or supply is too low). Deterioration in the quality and safety of services, or in customers’ satisfaction with taxi services, could indicate operators are trying to lower costs to match lower fare revenue because demand has fallen (or supply is too high).

Some performance measures of taxis in Sydney are available through the network standards and KPIs set by Transport for NSW and reported against by the taxi networks. However, these only apply to booked trips, which constitute a minority of all trips. The Taverner Research survey confirmed that only around a quarter of trips are booked through a taxi company.34 To date there has been no regular survey of customer satisfaction in NSW. IPART uses information on customer complaints and compliments from the Customer Feedback Management System operated by Transport for NSW as an indicator of service standards in our fare reviews, but we have recommended adoption of regular customer surveys as a better measure.

The Taverner Research survey sought to get a better understanding of the performance of taxi services from customers’ perspective. In general, 65% of people who had taken a taxi in the past 6 months were satisfied with the overall performance of the service. Customers were more dissatisfied with fares than with other aspects of taxi services. 36% of passengers said that they were dissatisfied with fares, whereas only 10% to 16% were dissatisfied with other aspects (such as waiting time, knowledge of Sydney and route taken). The highest rated aspects of the journey were the directness of the route (59% were

33 Taverner Research, Survey of Taxi Use in Sydney, November 2012. 34 Data from Taverner Research, Survey of Taxi Use in Sydney, November 2012, analysed by IPART.



very satisfied or satisfied) with driver courtesy and helpfulness somewhat lower (48%) but still well above the fare charged (26%).35

Most people felt that the quality of taxi services has not changed since last year. The main exception was for fares. Over half of taxi users surveyed (55%) said that the cost of fares is worse than it was last year, while only 6% considered this had improved. The next worse trend was in journey times with 16% saying this was worse against 8% saying it was better. Driver behaviour and the time taken to make a booking both attracted slightly more ratings of ‘better’ than of ‘worse’.36

The demand for new taxi-cab licences

The demand for new taxi licences is driven by factors such as the income a potential taxi operator expects to earn, the existence of other employment opportunities (including driving but not operating a taxi) and their expected incomes, and the availability and price of taxi licences and leases on the secondary market.

The viability and sustainability of the taxi-cab industry

Our approach for deciding on the number of licences to be released needs to consider the impacts on drivers and operators, as well as those on existing licence owners (discussed above) so that the industry continues to be viable in both the short and the longer term.

Any other matters considered relevant, having regard to the objective of ensuring improved taxi-cab services

Other matters can be considered in determining the number of licences to be released, provided they relate to improving taxi services for the community. We note that previous reviews have not considered additional matters. We have not identified any other matters for consideration.

3.3 Approach used to determine number of new annual licences to be released to date

In January 2010, immediately after the 2009 amendments came into effect, Transport for NSW released 100 new annual licences as an interim measure. Later that year, and in March 2011 and 2012, it made annual determinations of new annual licences to be released each year, in line with the amendments.

35 Taverner Research, Survey of Taxi Use in Sydney, November 2012, pp 46-47. 36 Taverner Research, Survey of Taxi Use in Sydney, November 2012, pp 59-60.



For each of those previous determinations, PricewaterhouseCoopers (PwC) provided advice to Transport for NSW. PwC’s recommendations were based on a formula based on available information – the Sydney Taxi Growth Model (STGM). In our issues paper, we discussed the STGM in some detail. It incorporates 9 components, which use a range of available data which reflect the factors listed in the legislation, discussed above.

Table 3.1 shows the numbers of licences released each year since 2010:

Table 3.1 Licences released each year since 2010

Year New annual licences (growth)

Replacement annual licences

2010 (interim) 100 0 2010 167 149 2011 197 30 2012 0a 57 Total 464 236

a The STGM returned a negative value in 2012. Source: Transport for NSW website. PwC, Annual taxi licence release 2012/13 – Final Report, 15 March 2012,

3.4 Success of this approach in meeting the amendment’s objectives to date

Now that the 2009 amendments have been in effect for 3 years, we can assess how successful they have been in achieving their objectives. Our view is that their success to date has been limited. The supply of taxis has grown by about 8% over these 3 years,37 which is below PwC’s estimate of long-term growth in demand of 3% to 5% per year.38 The number of new taxi licences has not been sufficient to make taxi market entry significantly more affordable, or place downward pressure on taxi fares.

3.4.1 We consider that the objectives of the legislative amendments have not been met

As Figure 3.1 shows, the cost of obtaining a licence for a taxi operator has changed little since 2009. Whether purchasing or leasing a licence, the cost remains high. Over the same period, taxi fares have continued to rise by more than the change in the Consumer Price Index or the change in the Wage Price Index.

37 Information provided by Transport for NSW on number of Sydney taxi licences at 1 January

2010 and 1 January 2013. 38 PwC, New South Wales Transport and Infrastructure – Analysis on release of new annual taxi licences

from 1 July 2010, March 2010, p 4.



Figure 3.1 Cost of obtaining a licence

Data source: Transport for NSW; PricewaterhouseCoopers, Benefit/Cost Assessment of Options for Reform of Taxi Licencing, September 2005.

In addition, the lack of change in the cost of leasing or purchasing a licence indicates there has been essentially no impact on existing licence holders. This suggests that the need to avoid unreasonable impacts on licence holders has outweighed (rather than been balanced with) the need for more affordable entry into the taxi market.

3.4.2 We have decided not to use the Sydney Taxi Growth Model to estimate the number of new licences needed

Our draft decision was to use a different approach to the STGM to estimate the number of new licences needed. This section discusses the feedback we received on our draft decision from submissions and the reasons why our final decision is not to use the STGM.

Our draft decision on the Sydney Taxi Growth Model

In our issues paper we indicated that we had concerns that the STGM was not a suitable method for calculating taxi licence release numbers and we sought feedback on whether there were any advantages in using a modified version of it.



In its submission on the issues paper, the NSW Taxi Council argued that most of the components in the model are required to comply with the requirements of the Act, and proposed modifications to overcome some of the concerns about data.39 However, as we noted in our draft report, the requirements of the Act do not have to be incorporated into a single formula, and in fact including components that are not specifically demand-related in a formula has helped to preserve the existing balance between supply and demand for taxis in Sydney, by acting as an in-built brake on the number of growth licences released. For example, when taxi service levels begin to improve, or the transfer values of existing licences start to decline, the STGM recommends fewer new taxi licences be released.

One submission argued that the STGM should be run in parallel with a new model as a validation exercise.40 However, this would only be worthwhile if we thought that the STGM was a valid method.

In our draft report, we decided that ‘given our own concerns about the STGM, and the lack of support or compelling arguments for maintaining it in a modified form, our draft decision is that a new approach is warranted.’41

Submissions on the draft report supporting the STGM

The NSW Taxi Council submitted that we should use the STGM for this review, with improvements as suggested by the NSW Taxi Council’s submission to the issues paper.42 St George Cabs reiterated that the STGM should be used in parallel to validate IPART’s recommendations.43

The NSW Taxi Council’s submission states that the STGM is simpler, more transparent, more understandable and predictable than IPART’s proposed approach, takes account of real-world performance and ensures the supply of taxis responds closely to growth in passenger demand. The submission also states that ‘the only justification that IPART has provided for rejecting the STGM is that it has not produced rapid enough change.’44

39 NSW Taxi Council submission, 9 November 2012, p 12. 40 St George Cabs submission, 9 November 2012, p 3. 41 IPART, Annual taxi licence release for Sydney 2013/14 – Draft Report, December 2012, p 21. 42 NSW Taxi Council submission, 21 January 2013, p 37. 43 St George Cabs submission, 21 January 2013. 44 NSW Taxi Council submission, 21 January 2013, p 35.



Our response to submissions supporting the STGM

While we did not reproduce the arguments in our draft report, our issues paper did discuss our concerns with the STGM approach.45 In summary, the STGM is not, despite its name, a model of the demand for taxi services, but an index constructed from available data, some of which are related to demand and some of which are not. The factors in the STGM which are directly related to demand are not included based on their quantitative relationship to demand, but in an arbitrary fashion. Some of these problems were acknowledged by PWC when they developed the STGM, and the STGM was proposed as a ‘next best’ solution that used objective inputs, was simple and transparent and enabled a ‘balanced consideration of factors’.46

While the STGM is certainly simple and transparent, during the time it was in use PwC and others expressed concerns about the data inputs,47 and it proved not particularly stable or predictable, returning a negative number in the third year it was used.48

The NSW Taxi Council acknowledged these data and stability shortcomings of the STGM in its submission on our issues paper, proposing improvements.49 Some of the NSW Taxi Council’s suggested changes (such as using longer term averages for proxy measures of demand) are consistent with the approach we have taken to estimating demand from external factors. (Chapter 4 discusses our approach to estimating demand from external factors in more detail).

However, we consider that the NSW Taxi Council’s proposed improvements do not address a fundamental problem with the STGM: that the inclusion of factors intended to capture measures of changes to ‘industry viability’ and ‘performance of existing services’ preserve the existing relationship between supply and demand for taxis. The NSW Taxi Council argues that we are required by our terms of reference to develop a formula that includes all the factors in the terms of reference (which in turn come from the Passenger Transport Act).50 We agree that we must meet our terms of reference; however, we do not agree that considering factors requires them to be included in a formula.

45 IPART, Annual taxi licence release for Sydney 2013/14 - Issues Paper, October 2012. 46 PwC, New South Wales Transport and Infrastructure – Analysis on release of new annual taxi licences

from 1 July 2010, March 2010, pp 12-19. 47 PwC, Annual taxi licence release 2012/13: Final Report, March 2012, p18 (network bookings not

representative of overall passenger demand), p 20 (errors in reporting of network bookings data).

48 Ibid, p 5. 49 NSW Taxi Council submission, 9 November 2012, p 11. 50 NSW Taxi Council submission, 21 January 2013, p 35.



Submissions on the draft report proposing a different approach and our response

Several submissions supported releasing an unlimited number of licences at a fixed price below the current market price, but only to operator/drivers.51 The submission from a former Victorian Taxi Inquiry commissioner noted that this approach had also been recommended by the Victorian Taxi Inquiry. His submission continued ‘although IPART may be constrained by its terms of reference, it is suggested that consideration again be given to the possible adoption of a price-based rationing system in NSW.’52

Our response to submissions proposing alternative approaches

As we noted in our issues paper, the legislation that governs the annual licence determination process does not permit this approach. A decision was made by Government when the amendments were made to the legislation in 2009 that release of licences should be determined by quantity rather than by price. Therefore our terms of reference ask us to recommend a fixed number of new licences, rather than recommend a price for an unlimited number of licences.

51 E Mollenhauer submission, 21 January 2013, p 2; NSW Taxi Drivers’ Association (NSWTDA)

submission, 21 January 2013, p 3; Australian Taxi Drivers’ Association (ATDA) submission, 23 January 2013.

52 D Cousins submission, 15 January 2013, pp 2-3.

4 The demand for taxi services



Under Section 32C(3) of the Passenger Transport Act 1990, regard must be had to the likely passenger demand for taxi-cab services, including latent demand, and the performance of existing taxi-cab services.

We firstly considered the external factors that are likely to increase the demand for taxi services, such as growth in population, household and business income, economic activity and tourism. Growth in these factors is likely to increase the size of the potential market of taxi users, and increase the frequency with which people use taxi services.

By examining trends in NSW state final demand, population growth and airport passenger numbers, we consider that demand for taxi services is likely to grow by around 2.5% per year over the longer term. In our view, at least an additional 140 new unrestricted licences should be issued each year to keep pace with this growth. We recommend that this be set as the minimum number of licences that will be issued each year over the next 5 years. These are the same as the decisions we made in our draft report.

We also considered the demand that would be generated by changes to the number of taxis on the road, and the price of taxi services. In particular: if there were more taxis on the road, additional trips would be taken because

passengers would not have to wait as long to catch a taxi if fares were to fall, additional trips will be taken because it costs less to use

taxi services.

This is known as latent demand, because some people are not currently making these trips, but would make them if waiting time or prices were reduced. As part of our assessment of latent demand we looked at the performance of existing taxi services by surveying Sydney residents to measure the current time taken to catch a taxi. Chapter 5 provides more information on the impact of latent demand on our recommendations.



4.1 How changes in demand relate to the change in licence numbers

Our terms of reference require us to consider the need for the supply of taxis to respond closely to the growth in passenger demand. We must also consider the need to reduce barriers to entry and place downward pressure on fares.

At a minimum, the number of taxi licences needs to increase in line with the demand due to external factors to ensure that lease costs do not increase in real terms (that is, after adjusting for inflation). But additional licences will be required for lease costs to fall in order to make it more affordable to provide taxi services and to place downward pressure on fares.

Additional licences will also be needed to reduce waiting time for customers. Latent demand would convert to actual demand when these new trips are taken because passengers will not have to wait as long to catch taxis.

4.2 Estimating changes in demand due to external factors

Taxi use increases as external factors expand the size of the taxi market. Growth in population, household and business income, economic activity, and tourism are all likely to increase demand for taxis in Sydney. We have used the following indicators to measure the changes in these drivers of demand: population growth in Sydney

state final demand – as an indicator of general economic conditions, capturing income growth from local households and businesses

airport passenger numbers – as an indicator of tourist/business activity.

Table 4.1 Change in external sources of demand

2008 2009 2010 2011 2012 AAGR

Population growtha 1.6%d State final demandb 4.7% -1.9% 4.0% 2.4% 3.4% 2.5% Airport passenger numbersc 5.7% -1.1% 6.5% 4.3% 0.1% 3.1%

a Actual population growth is not published annually. b June quarter figures reported. Percentages are calculated as June quarter for the relevant year on June quarter for the previous year. c Figures are for financial year to June of the relevant year. d Average annual growth rate from 2006-2011 published by ABS. Source: State final demand - 5206.0 Australian National Accounts: National Income, Expenditure and Product - Table 21. State Final Demand, Summary Components by State: Chain volume measures (Seasonally adjusted). Population growth - ABS, 31010DO001_201203 Australian Demographic Statistics, Mar 2012, http://www.abs.gov.au/ausstats/[email protected]/mf/3101.0, Airport passenger numbers - Department of Infrastructure and Transport, Bureau of Transport, Infrastructure and Regional Economics, Air passenger movements through capital and non-capital city airports to 2030–31, Report 133, November 2012, p 25.

http://www.abs.gov.au/ausstats/[email protected]/mf/3101.0



Table 4.1 shows that the recent trends indicate that demand is likely to grow by 2.5% per year on average due to external factors.

We recommend for this review and the next 4 annual reviews of taxi licences that the minimum increase in the number of licences is 2.5%, reflecting a longer term estimate of growth in demand for taxi services resulting from external factors. For simplicity and certainty we consider that the 2.5% should be applied to the current number of taxi licences as at 1 January 2013, so that a minimum of 140 new licences are released each year. We expect that increasing the number of taxi licences by this amount each year in line with the increase in demand from external sources would mean that lease costs remain unchanged in real terms.53 As mentioned previously, more licences than this will need to be released for lease costs to fall to improve affordability into the industry, and to improve waiting times for customers to convert latent demand to actual demand.

Submissions on the draft report generally did not support our view of demand. The sections below set out our reasoning, the comments made in submissions, and our response to them.

4.2.1 Indicators of external sources of demand

In submissions on the issues paper, stakeholders suggested that we should look at a range of indicators of demand for taxi services:

NSW GDP, state final demand54 employment rate, movement in wages, commercial tenancy occupancy rates55 population growth56

airport passenger numbers57 measures of economic uncertainty – rates of corporate cost cutting and

consumer spending.58

We agree that many of these factors will influence the amount of taxi use in Sydney. Regression analysis is a statistical tool that could be used to determine the strength of the relationship between these and other variables and taxi use. The outputs could then be used to estimate the change in demand from one period to another. While we consider that this would be the best way to measure

53 That is, assuming fares and all other costs of providing taxi services increase by the rate of

inflation, lease costs would also continue to rise by the rate of inflation. 54 N O’Brien submission, 9 November 2012, p 1; NSW Taxi Council submission, 9 November 2012,

p 3. 55 E Atra submission, 5 November 2012. 56 NSW Taxi Council submission, 9 November 2012, p 4; Sydney Metro Taxi Fleet submission,

November 2012; D Lipski submission, 9 November 2012, p 5; E Atra submission, 5 November 2012.

57 NSW Taxi Council submission, 9 November 2012, p 4. 58 NSWTDA submission, 9 November 2012, p 2.



the change in demand from these factors, we do not have enough historical information of the number of trips taken in Sydney to undertake this analysis.

Therefore we have decided to focus on the same 3 indicators of external sources of demand that were used previously by PwC: state final demand, Sydney population size and Sydney airport passenger numbers. Many of the other measures suggested in submissions are closely related to these. Continuing to use these factors was generally supported in submissions on our issues paper.59

Why state final demand?

State final demand reflects the relationship between income and spending on goods and services, including taxi services. It includes government and household final consumption expenditure. State final demand has been used in previous years.60 The NSW Taxi Council supported PwC’s findings from previous reviews that state final demand was the most relevant measure of economic activity.61

Why population growth?

Population has also been used in the past to forecast the demand for taxi services. The NSW Taxi Council submitted that while it is appropriate to consider population growth, ‘there is no conclusive evidence of a strong correlation in the increased demand for taxis and population increase’.62

Why passenger numbers at Sydney Airport?

In previous reviews, the PwC considered passenger numbers at Sydney airport as an indication of the likely change in the number of taxi journeys to and from Sydney airport. The NSW Taxi Council supports using airport passenger numbers because it is a significant source of demand for taxis.63

We consider that passenger numbers at Sydney airport are an indicator not only of the trips likely to be taken at Sydney airport, but also the growth in the size of the taxi market reflecting tourism and business growth. Increases in airport passengers suggest greater tourist and business travel activity.

59 We note that PwC used network bookings to measure the change in demand. We have not

included network bookings because we do not consider them to be a good indicator of demand for taxi services.

60 PwC, Annual taxi licence release 2012/13 – Final Report, 15 March 2012. 61 NSW Taxi Council submission, 9 November 2012, p 7. 62 Ibid, p 4. 63 Ibid, p 4.



4.2.2 Using historical trends rather than annual figures

Annual measures of indicators of external sources of demand were used in the previous reviews of licence numbers. However, state final demand and airport passenger numbers can fluctuate significantly from year to year. For example, Figure 4.1 shows the volatility in state final demand over time. Therefore using annual figures to predict demand will mean that new licence numbers will vary significantly from year to year, creating uncertainty in the industry without more accurately matching demand and supply.

Figure 4.1 ABS – Annual percentage change in state final demand

Note: Percentage changes are calculated as June quarter figure for the relevant year on the June quarter figure for the previous year. Data source: ABS : 5206.0 Australian National Accounts: National Income, Expenditure and Product - Table 21. State Final Demand, Summary Components by State: Chain volume measures.

In tis submission on our issues paper, the NSW Taxi Council submitted that historical trends rather than annual numbers should be used to produce stable and predictable results. It considered that we should use a 3-year compound annual growth rates for each indicator.64 We agree that trends should be used to reduce volatility and provide more certainty for the industry. For this review we have considered the last 5 years of data.

We also note that in the past, PwC focused on population forecasts to have a forward-looking measure in the STGM.65 For consistency with our other measures, we have considered historical trends as an indicator of likely population growth.

64 NSW Taxi Council submission, 9 November 2012, p 3. 65 PwC, Annual taxi licence release 2012/13 – Final Report, 15 March 2012, p 21.

-4%

-2%

0%

2%

4%

6%

8%

June 1987 June 1992 June 1997 June 2002 June 2007 June 2012

SFD seasonally adjusted (annual % change)5 per. Mov. Avg. (SFD seasonally adjusted (annual % change))



State final demand

Over the past 5 years annual growth in state final demand has averaged 2.5%. Over the longer term, the rolling 5-year average is in the range of 2% to 4%. The NSW Taxi Council suggested using a 3-year Compound Annual Growth Rate for state final demand, which the NSW Taxi Council calculated at 2.92%,66 but we prefer the 5-year annual average growth rate as a more stable long-term indicator.

Population growth

The most recent historical data from the ABS (2010/11) shows that population has grown on average by 1.6% per year over the past five years.67 These numbers are slightly higher than the Department of Planning forecast. Using the Department of Planning forecast, the NSW Taxi Council recommend using an increase of 1.22% in population growth based on a 3-year compound annual growth rate,68 but we have preferred the 5-year historical data for consistency with our other measures, and for more long-term stability.

Airport passenger numbers

Data from the Bureau of Infrastructure, Transport and Regional Economics (BITRE) shows an average increase of 4.6% per year since 1991/92.69 It forecasts an average 3.6% per year over the next 30 years.70 The NSW Taxi Council have noted that the most recent 3 years of data shows an average compound annual growth rate of airport passenger numbers of 2.11%.71 We have preferred the 5-year historical annual average of 3.1% for consistency with our other measures, and for more long-term stability.

66 NSW Taxi Council submission, 9 November 2012, p 7. 67 ABS, 31010DO001_201203 Australian Demographic Statistics, Mar 2012,

http://www.abs.gov.au/ausstats/[email protected]/mf/3101.0. 68 NSW Taxi Council submission, November 2012, p 8. 69 Department of Infrastructure and Transport, Bureau of Transport, Infrastructure and Regional

Economics, Air passenger movements through capital and non-capital city airports to 2030-31, Report 133, November 2012, p 25.

70 Department of Infrastructure and Transport, Bureau of Transport, Infrastructure and Regional Economics, Air passenger movements through capital and non-capital city airports to 2030-31, Report 133, November 2012, p 25.

71 NSW Taxi Council submission, November 2012, p 8.

http://www.abs.gov.au/ausstats/[email protected]/mf/3101.0



Comments on demand from external factors from submissions

Most submissions took the view that demand is stagnant or declining, citing various reasons, including the global economic downturn, competition from hire cars, technology changing work practices, businesses restricting staff use of taxis, the strong Australian dollar reducing tourism, and the experience of cab drivers on the streets.72 Two submissions73 pointed to the declining trend in taxi bookings (an input into the STGM) as indicators of declining demand.

The NSW Taxi Council submitted that we did not explain the weighting of each indicator, nor how we obtained the figure of 2.5%. The NSW Taxi Council also opposed our longer-term recommendation that each of the next 4 reviews release at least 140 licences, describing it as a fixed growth policy that was contrary to the intention of the amendments to the Passenger Transport Act that set up an annual review process.74

Former commissioner of the Victorian Taxi Inquiry David Cousins questioned ‘the assumption that the so-called base level of demand can be adequately determined by a formula approach’, submitting that IPART had ‘strongly and effectively criticised the use of formulas to estimate demand in its draft report’.75

Response to views on demand growth from external factors raised in submissions

Our estimate of long-term annual demand growth from external factors is not based on a formula, but a judgement following examination of three relevant indicators. The measures of demand used in the STGM were quite volatile, so we preferred an approach that gave a more stable result over time. However, it is not a ‘fixed growth policy’ as there will continue to be an annual review and determination.

4.3 Estimating latent demand

Latent demand represents the additional taxi trips that would be taken if waiting time for taxi services fell, or if prices were reduced. New licences, in addition to the licences for growth in external sources of demand, will be required in order to reduce waiting times for customers.

72 Anonymous submission (W13/2), 3 January 2013; H Batth submission, 21 January 2013; P

Nicolopoulos submission, 21 January 2013; T Fathinia submission, 14 January 2013; D Gill submission, 10 January 2013; N O’Brien submission, 21 January 2013; M Mikhail submission, 29 December 2012; Anonymous submission (W13/48), 16 January 2013; M Gordon submission, 18 January 2013; M Burrage submission, 12 January 2013; NSWTDA submission, 21 January 2013; RSL Cabs submission, 18 January 2013; S Porcaro submission, 21 January 2013; A Yazdabadi submission, 21 January 2013; ATDA submission, 23 January 2013.

73 NSW Taxi Council submission, 21 January 2013, p 18; ATDA submission, 23 January 2013, p 5. 74 NSW Taxi Council submission, 21 January 2013, p 18. 75 D Cousins submission, 15 January 2013, p 2.



As mentioned in chapter 3, we commissioned Taverner Research to conduct a survey of 2000 Sydney residents to help us understand the behaviour of taxi users. The survey results informed our estimates of responsiveness to price and the time taken to catch a taxi. In particular, we were able to collect information on: the average time that people are currently waiting to catch a taxi the main reasons for catching and not catching taxis, including the importance

of being able to catch a taxi quickly satisfaction with different features of taxi services.

4.3.1 How changes in fares affect the number of trips taken

Price elasticity measures the response in demand to fare changes - that is, the increase in the number of trips when prices fall, and the decrease in the number of trips when prices rise.

Our survey results indicate that price is one of the most important factors when considering whether to catch a taxi. For example: of the 542 people who thought about catching a taxi in the last 6 months, but

in the end decided not to, 67% reported that the reason they decided not to was because it was too expensive76

of the 671 people who didn’t consider catching a taxi in the last 6 months, 54% said one of the reasons for not using a taxi was because it is too expensive.77

Price elasticity for taxi services is likely to vary at different times of the day, because responsiveness to prices depends largely on the availability of substitutes. For example, at night, when public transport services are less frequent, people are likely to be less sensitive to price – even if the price rose significantly they would still catch taxis because there are no other options available.

This is consistent with the results from our taxi survey which showed that people who caught taxis at night were the most dissatisfied with fare levels.78 The fact that these customers were less satisfied with fare levels but took the journey anyway suggests that they are less price-sensitive. Our survey also showed around 20% of people catching a taxi after 10 pm did so because it was their last resort, compared to less than 10% at all other times.79

76 Taverner Research, Survey of Taxi Use in Sydney, November 2012, p 36. 77 Taverner Research, Survey of Taxi Use in Sydney, November 2012, p 66. 78 Taverner Research, Survey of Taxi Use in Sydney, November 2012, p 55. 79 Taverner Research, Survey of Taxi Use in Sydney, November 2012, p 45.



We have assumed an average price elasticity value across all shifts of -0.8, which means that for every 1% fall in prices, the number of new trips increases by 0.8%. It also means that the total revenue from all taxi fares would fall by 0.2%. A value of elasticity of -0.8 is comparable with the Victorian Taxi Inquiry’s final report, which suggests an elasticity of around -1 for Melbourne.80 Our estimate is also in the range reported by Booz Allen Hamilton, in a report for IPART in 2003. It noted that the majority of international studies reported a price elasticity of demand of -0.2 to -1.81

4.3.2 How changes in the time taken to catch a taxi affect the number of trips taken

Waiting time elasticity measures the response in demand to changes in waiting times - that is, the increase in the number of trips when waiting time falls or the decrease in the number of trips when waiting time increases.

The results from our survey suggest that the amount of time that it takes to catch a taxi is important to the decision to use a taxi. Around 30% of taxi users in our survey said that their main reason for catching a taxi is because it was quicker (or more direct) than other options.82 Waiting time is also important in people’s decision not to catch a taxi. For example:

Of the 162 passengers surveyed who tried to catch a taxi but in the end did not: – 37% gave up because the wait at the taxi rank was too long – 35% gave up because they didn’t wait long enough to see any vacant taxis

driving by.83

Similarly, of the 542 passengers surveyed who thought about catching a taxi but in the end decided not to: – 12% chose not to catch a taxi because they were not sure if it would turn up

in a reasonable period of time if they booked a taxi – 14% didn’t wait long enough to see any vacant taxis driving by – 7% thought that the wait at the taxi rank was too long.84

80 The Victorian Taxi Inquiry estimated that on average across all customer segments, a 10%

reduction in fares would increase demand by 10.4%. Victorian Taxi Inquiry, Customers First: Service, Safety, Choice, May 2012, pg 440, http://www.taxiindustryinquiry.vic.gov.au/__data/assets/pdf_file/0004/67873/09-Part-E-20120626.pdf; an elasticity of -1 implies lower revenue from falling fares is completely offset by an increase in demand.

81 Booz Allen Hamilton, Appraisal of taxi fare structure issues, July 2003, p 10. 82 Taverner Research, Survey of Taxi Use in Sydney, November 2012, p 44. 83 Taverner Research, Survey of Taxi Use in Sydney, November 2012, p 34. 84 Taverner Research, Survey of Taxi Use in Sydney, November 2012, p 36.

http://www.taxiindustryinquiry.vic.gov.au/__data/assets/pdf_file/0004/67873/09-Part-E-20120626.pdf




Of the 671 passengers surveyed who didn’t even consider catching a cab in the last 6 months, over 25% said the reason why they didn’t consider catching a taxi was because they thought the waiting times would be too long, or they were worried that a taxi would not show up after they booked one.85

The results from our taxi survey indicate that the current average waiting time for a taxi is around 8.5 minutes.86 However, this varies significantly by time of day, and also depends on how the taxi was obtained.

Figure 4.2 shows that waiting times were longer on Friday or Saturday in the evening (6pm to before 10pm) or overnight (10pm to before daybreak) than in the afternoon (midday to before 6pm) or morning (daybreak to before midday). For Monday to Thursday, the longest waiting times were in the afternoon and waiting times overnight were mostly under 5 minutes.

85 Taverner Research, Survey of Taxi Use in Sydney, November 2012, pp 65-66. 86 Data from Taverner Research, Survey of Taxi Use in Sydney, November 2012 and the taxi

industry model.



Figure 4.2 Waiting time by time of day within day of week for trips taken from a rank, hailed, and booked for the ‘next available taxi’

Note: This figure is different from Figure 4.2 in our draft report, which also included waiting times for passengers who had booked their trip for a particular time. Data source: Taverner Research, survey data.

Figure 4.3 shows that passengers who booked a taxi waited longer than customers who hailed taxis or boarded at a rank (although this might not have included the time taken to reach the rank). Booking the next available taxi was substantially more likely to involve a wait of 10 minutes or more and for a few (3%) the wait was 40 minutes or more.

48%

42%

52%

71%

52%

48%

29%

36%

47%

46%

34%

35%

28%

13%

30%

29%

41%

35%

29%

25%

16%

17%

18%

16%

14%

21%

25%

23%

24%

25%

2%

5%

2%

5%

3%

5%

6%

4%

0% 20% 40% 60% 80% 100%

morning n=(149)

afternoon n=(155)

evening n=(96)

overnight n=(45)

morning n=(44)

afternoon n=(63)

evening n=(133)

overnight n=(171)

morning n=(17)

afternoon n=(24)

Mon

day

to T

hurs

day

Frid

ay o

r Sat

urda

ySu

nday

Less than 5 minutes 5 to less than 10 minutes10 to less than 20 minutes More than 20 minutes



Figure 4.3 Waiting time by how taxi obtained

Data source: Taverner Research, survey data.

4.4 Measuring the overall change in demand based on taxi trip data

Many stakeholders recommended obtaining and using taxi trip data to measure changes in demand. We agree that taxi trip data is vital to good regulation because it will improve our understanding of how the number of journeys responds to our recommendations on fares and taxi licence numbers.

Taxi trip data will incorporate demand that is generated due to both external factors (changes in population, income, tourism and business activity), and changes in demand due to waiting times and fares. For example, if fares were to increase significantly in one year, after allowing for the change in demand due to external factors, the remaining change in the demand would largely represent the price elasticity of demand.

4.4.1 A comprehensive regulator data set direct to the regulator

In a report prepared for the Victorian Taxi Industry Inquiry, Rhumb Consulting recommended mandating a regulator dataset to be transmitted from operators through metering technology directly to the regulator. In our view, implementing such an arrangement is in the best interests of the industry, passengers and regulators. Much of the data needed is currently available but lack of agreement over what data is required and the format for presentation means it is costly to provide quality data in a timely way.87

87 Rhumb Consulting, Taxi Industry Inquiry Information System Review – Summary of Key

Recommendations, 9 December 2011, p 6.

58%

45%

9%

29%

36%

38%

12%

17%

40%

1%

2%

10% 3%

0% 20% 40% 60% 80% 100%

rank (n=366)

hail (n=376)

booked next available taxi (n=173)

Less than 5 minutes 5 to less than 10 minutes 10 to less than 20 minutes20 to less than 40 minutes Over 40 minutes



Comments in submissions on the regulator data set

Few submissions commented on the data collection recommendations in our draft report. Two submissions88 supported the development of a regulator data set, while one thought the requirement ‘may be onerous to networks and not really necessary’.89

Recommendation

1 Transport for NSW should investigate the cost and feasibility of mandating a regulator data set.

4.4.2 An interim solution – sample of meter data from Sydney taxi networks

As an interim measure until a comprehensive data set such as that described above is mandated, it makes sense to make existing data available for this purpose. Therefore, we consider that Roads and Maritime Services should require authorised networks to provide IPART with meter swing data (data on actual taxi trips) already captured and stored by taxi networks for a specified sample of weeks on an annual basis commencing in 2013.90

PwC has previously noted the need for access to this ‘meter swing’ data. Networks and the NSW Taxi Council argued against requiring them to provide this information to TfNSW/PwC on the basis that:

historical data is unlikely to be available

data might not be accurate or reliable

taxi drivers strategically activate/deactivate meters in order to manipulate network dispatch systems or avoid putting work through the meter

in busy times network dispatch systems might not collect meter swing data at all (they devote limited system capabilities to other tasks).

88 D Cousins submission, 15 January 2013, p 4; ATDA submission, 23 January 2013, p 5. 89 T Bradley submission, 22 January 2013. 90 The Passenger Transport Regulation 2007 (s182) requires networks to report on performance

‘relating to the provision of taxi cab services’ to Roads and Maritime Services (RMS). The regulation provides for reports to be requested at any time and for RMS to specify the content, format and timeframe in which they must be provided (provided such a request is reasonable).



In order to minimise costs we consider that only a sample of data needs to be provided. We are seeking data on: meter activations (number of paid trips) occupancy (occupied time per hour on the road)

occupied km per km on the road number of trips booked through a network shift length (start and finish time)

distance per paid trip, time taken per paid trip.

This data should be compiled: by shift on a per taxi basis, for all taxis affiliated with the network

for the first full week in each of March, June, September and December (Monday to Sunday), commencing in June 2013.

The networks should provide this data in an Excel template common to all networks. We will liaise with TfNSW in order to ensure that the template meets our needs and where possible is compatible with network systems. If networks would like to suggest a format for this report they should do so.

We will also seek downloads of meter data directly from operators on a voluntary basis, which would provide information on fares charged that we cannot get via networks. We would need to contact operators directly and ask them to collect this information for the sample weeks.

Comments in submissions on interim data collection

The only comment we received on the interim proposal was from David Cousins, former commissioner of the Victorian Taxi Inquiry. He submitted that there might not be much benefit in seeking sample data in the absence of a comprehensive regulator dataset. He also thought that networks were not the best source of data.91

However, we continue to be of the view that collecting sample data until the regulator data set can be established would facilitate better regulatory decision-making.

Recommendation

2 Roads and Maritime Services should require authorised Sydney taxi networks to provide reports on utilisation of all affiliated taxis for a sample of 4 weeks per year commencing in June 2013.

91 D Cousins submission, 15 January 2013, p 4.

5 The number and mix of licences to be released



In determining the number of annual licences to be issued, we considered the likely passenger demand for taxi-cab services, including latent demand. As we indicated in the previous chapter, we consider that licence numbers should increase by a minimum of 2.5% in 2013 and each year for the next 4 years to account for the growth in demand caused by external factors including population, income, business activity and tourism activity.

However, if the growth in licence numbers simply keeps pace with the growth in demand from external factors, lease costs are likely to be maintained in real terms.92 In order to achieve the objective of more affordable entry into the taxi market, more licences need to be released than the minimum 2.5% increase.

On the other hand, if the additional licence numbers greatly exceed the growth in demand from external factors, lease costs will fall substantially, which means a bigger impact on the incomes of existing owners of perpetual licences. Our terms of reference require us to balance the 2 objectives of providing more affordable leases for operators and avoiding unreasonable impacts on licence holders.

Determining how many additional licences to release over and above the base level of demand cannot be accomplished by a simple formula. The taxi industry consists of complex interactions between supply of taxis, demand for taxi services, waiting time and fares. Different classes of licence (such as Peak Availability Licences (PALs)) also have a different impact on the way the market works.

We used a model of the taxi industry developed by the CIE to help us to understand the likely outcomes of adding different numbers and types of licences to the Sydney taxi market,93 and ultimately made a judgement about the number and mix of taxi licences that would maximise the benefits and minimise the impacts of releasing new licences, thereby improving taxi service performance and enhancing the viability and sustainability of the industry.

92 That is, if fares and other costs of providing taxi services increase in line with inflation, lease

costs will increase in line with inflation too. 93 More detail on the model is contained in Appendix C.



Our decision is to recommend releasing 160 new unrestricted licences and 200 new PALs. We also recommend releasing 70 unrestricted and 30 PAL replacement licences, and that an adjustment for Wheelchair Accessible Taxi licences should be made. We consider that this combination meets the requirement to consider demand and latent demand: 140 unrestricted licences fits closely with 2.5% underlying growth in demand

across all shifts.

A further 20 unrestricted licences and 200 PALs will help to match the demand for taxis that varies throughout the week with the number of taxis on the road to meet it. Ideally, any new licences would be unrestricted, and operators would choose freely which times to put them on the road, but releasing more PALs will assist the industry to adapt to a more flexible way of meeting demand through enabling a different business model at a different price point.

The 200 PALs will be available throughout the week between midday and 5 am, so they will be able to address the demand and latent demand on a Friday and Saturday night. They are also more likely to be on the road during the 3pm and 3am changeover times for unrestricted taxis. This means that the number of PALs increases by 70% (not including the PALs for replacement).

Recommendation

3 Transport for NSW should offer for tender 160 new unrestricted annual licences and 200 new peak availability licences (PALs) in 2013/14.

This is slightly different from our draft recommendation, which was to release 140 new unrestricted and 250 new PALs. We are recommending more unrestricted and fewer PALs on the basis of providing more fully flexible licences to meet additional demand than we had previously recommended, while still providing enough PALs to facilitate a new business model.

We have also used the taxi industry model developed by the CIE to estimate the results of releasing this number and mix of licences and concluded that it provides an appropriate balance between the expected outcomes. That is, between positive outcomes for service performance (reduced waiting times and more taxi trips), driver income (taxi occupancy and driver hourly earnings), operator costs (lower annual licence costs) and impacts on existing licence-holders (lower income from perpetual licences).



Taking into account the additional 2.5% demand due to external factors and assuming fares stay at their current nominal levels and that all other costs of providing taxi services increase by the rate of inflation, we estimate the effects of this release to be:

6% increase in the total number of trips 1% increase in average occupancy per taxi (per hour) 10% reduction in lease income to owners of perpetual unrestricted licences

10% reduction in the annual cost of a licence or lease for operators 3% average reduction in waiting time for passengers 6% average reduction in waiting time for passengers on a Friday and Saturday

night.

For this review of licence numbers we have modelled outcomes assuming fares stay at current nominal levels. We are currently reviewing fares to apply from 1 July 2013 as part of our regular annual fare review. The issues paper for the fare review is available on our website. For the purposes of illustration for the licence review, we have modelled the different outcomes of the options for fares we have raised in the fare review issues paper. Section 5.3.6 shows the outcomes of our licence recommendations with different fare changes.

5.1 Balancing a more affordable means of entry into the taxi market with avoiding unreasonable impacts on existing licence-holders

Every licence released over the level of growth in demand from external factors would be expected to have an impact on the value of licence leases and thereby make entry into the taxi market more affordable.

The CIE’s taxi industry model can be used to estimate the market response to an increase in licence numbers. Changing fares in the model will also change the modelling results, as people change their taxi-catching behaviour in response to fares going up or down. Because we have only just commenced our fare review, we have decided to keep fares constant at their current nominal rates in the model.

We modelled different combinations of new licences to examine the likely impacts on lease costs, passenger waiting time, and change in the proportion of a shift that a taxi has a passenger. Figure 5.1 includes some of these combinations of new licences, and the impacts of each. The red box shows our recommendation on the combination of licences that should be released.



Figure 5.1 Licence release impacts

Data source: Taxi industry model.

We consider that a 10% reduction in licence lease costs is an appropriate balance between improved affordability for operators and reduced income for licence owners. We estimate that a 10% reduction in licence lease costs will result from 280 additional unrestricted licences or 480 PALs. There are other combinations of unrestricted licences and PALs that will have the same impact (with a total number of additional licences between 280 and 480).



We also consider that in order to meet the objectives of the legislation more licences will need to be released in future years. We recommend releasing 160 new unrestricted licences and 200 new PALs94 in 2013/14 in the context of a possible reduction in licence lease costs of around 25% over the next 5 years. However, in the event that we are given a referral to recommend licence numbers for 2014/15, we will review the context at that time.

5.1.1 Comments from submissions on the number of licences to be released

The majority of submissions expressed the view that the number of licences we recommended for release in the draft report was too many. Many stated or implied that no new licences should be released.95 Two submissions suggested issuing replacements for handed-back licences only.96

Many submissions argued that Sydney is already over-supplied with taxis and that if people can’t get taxis this is not due to under-supply of taxi licences, but other reasons, including: taxi drivers choosing where and when to drive, based on potential passenger

behaviour or perceived financial or personal risk to the driver (eg, avoiding Kings Cross and other entertainment precincts, avoiding congested areas, not taking bookings that drivers judge to be a no-show risk)97

not enough drivers for the existing cabs meaning operators can't get cabs on the road as much as they’d like98

too many No Stopping zones in the CBD and congestion problems around the CBD, entertainment precincts and the airport mean taxis can’t pick up would-be passengers99

infrastructure capacity constraints at the airport.100

94 Before the adjustments discussed in Chapter 6. The total numbers recommended for release are

225 unrestricted licences and 230 PALs. 95 M Mikhail submission, 29 December 2012; Anonymous submission (W13/2), 3 January 2013;

Anonymous submission (W13/3), 3 January 2013; E Atra submission, 6 January 2013; M Kiani submission, 8 January 2013; D Gill submission, 10 January 2013; Anonymous submission (W13/30), 11 January 2013; M Burrage submission, 12 January 2013; T Fathinia submission, 14 January 2013; S Guy submission, 16 January 2013; Anonymous submission (W13/48), 16 January 2013; I Pergamalis submission, 16 January 2013; G Koutsioukis submission, 17 January 2013; Anonymous submission (W13/64), 18 January 2013; NSWTDA submission, 21 January 2013; E Ieraci submission, 21 January 2013; A Yazdabadi submission, 21 January 2013; P Nicolopoulos submission, 21 January 2013; Anonymous submission (W13/90), 21 January 2013; T Bradley submission, 22 January 2013.

96 M Gordon submission, 18 January 2013, Anonymous submission (W13/67), 18 January 2013. 97 NSW Taxi Council submission, 21 January 2013, pp 10-11. 98 Anonymous submission (W13/3), 3 January 2013; S Guy submission, 16 January 2013;

NSWTDA submission, 21 January 2013, p 3; RSL Cabs submission, 18 January 2013; St George Cabs submission, 21 January 2013.

99 Manly Warringah Cabs submission, 9 January 2013; D Lipski submission, 21 January 2013, pp 12-17.

100 Taxi Council submission, 21 January 2013, p11; D Lipski submission, 21 January 2013, p 10.



Other submissions conceded that there might be times of the week or times of the year that are under-supplied with cabs, but argued similarly that this is not because of lack of licences but for structural or cultural reasons, or, alternatively, is not a problem: one submission expressed it as the taxi industry’s approach should be ‘enough cabs to service the public ALMOST all of the time’.101

Submissions suggested that we should look to increasing the utilisation of the existing fleet rather than adding more licences. Specific proposals included:

provide more taxi pick-up spots102 increase public transport at peak demand times for cabs103 fine taxi drivers for refusing short fares104

make owner/operators bail their cabs out for the shifts they don’t drive themselves105

require networks to have a specific number of cars on the road for special events106

look at ways to address the shift changeover problem.107

5.1.2 Response to comments on the number of licences to be released

Our own research showed that under-supply is only a problem at particular times of the week. Off-peak times (and the CBD and airport in particular) may well be over-serviced. However, we consider that ‘enough cabs to service the public almost all of the time’ is not the optimal situation for the taxi industry.

The taxi industry is one where demand varies within the day, across the week and throughout the year, but the supply (ie, the number of cabs) is fixed at any point in time. Taxis also have high fixed costs (ie, costs that are incurred whether or not the taxi is on the road, such as lease costs, network fees, registration and insurance), so the industry model where supply of taxis is able to match demand and all industry participants recover their costs is complex. We consider that the number and type of additional licences we have recommended will allow the industry to better match supply to demand.

101 D Petrovic submission, 21 January 2013. 102 D Lipski submission, 21 January 2013, pp 12 -17. 103 M Mikhail submission 29 December 2012; P Andersen submission, 13 January 2013; Legion

Cabs submission, 15 January 2013, p 5; St George Cabs submission, 21 January 2013. 104 R Bentley submission, 9 January 2013. 105 S Guy submission, 16 January 2013. 106 G Pavlis submission, 20 January 2013. 107 I Pergamalis submission, 16 January 2013.



We agree that inefficient utilisation of the taxi fleet is potentially an impediment to meeting demand, and we would encourage the taxi industry and other stakeholders to work to overcome any identified problems. However, our terms of reference ask us to recommend the number of licences that will ensure that the supply of taxis is adequate, and we consider that improvements to the balance between supply and demand can still be made by adding more licences, for any given level of fleet utilisation.108

5.2 Determining the mix of unrestricted and Peak Availability Licences (PALs)

Under our terms of reference, we can make recommendations on which class of licence should be released. Our final recommendation is to release an additional 200 PALs as well as 160 unrestricted licences.109 Ideally, licences would all be unrestricted, to allow the market to decide where and when to operate cabs, but in our view releasing some licences as PALs is a measure to help the taxi industry transition to a more flexible structure with a range of business models and operating patterns.

Our draft decision was to recommend slightly more PALs and slightly fewer unrestricted licences, but as noted above, for our final report we have decided to tip the balance more towards letting the market decide when to operate cabs, while still supporting the transition towards a mix of business models with PALs. We expect there to be less emphasis on PALs in future reviews.

5.2.1 Peak Availability Licences in the current fleet

Table 5.1 summarises the current fleet composition.

Table 5.1 Composition of taxi licences (1 January 2013)

Licence class Number Proportion of fleet

Unrestricted 4729 83.7% PALs 286 5.1% Fringe Area 14 0.2% WATs 618 10.9% Total 5647

Source: Transport for NSW.

108 We would also point out that a taxi fleet operating at maximum efficiency would not have 100%

of the available licences on the road at all times, or even at peak times. There will always be a percentage of cabs off the road for physical reasons (eg, maintenance, repairs) or operational reasons (eg, driver breaks, operator preference).

109 Before the adjustments discussed in Chapter 6. The total numbers recommended for release are 225 unrestricted licences and 230 PALs.



PALs can only be operated between 12 noon and 5 am, and are typically driven for a single shift (12 out of the possible 17 hours). This means that many of these taxis are on the road at 3pm and 3am changeover times, which helps address the shortage of taxis on the road at these times. Currently PALs make up about 5% of all licences in the Sydney market. Through the annual tender process, the annual licence costs for PALs have been about 75% of the price of an unrestricted licence.110

5.2.2 Releasing more PALs maximises improvements in performance for a given impact on licence holders

Our terms of reference require us to consider the performance of existing taxi-cab services. Therefore, we think that, for a given reduction in income from leases, we should seek to maximise performance improvement.

Survey results111 confirm that customers face the longest waiting times, and have the most trouble catching a taxi, on Friday and Saturday nights. Therefore the more taxis that are on the road during this period, the more tangible the service improvements are likely to be. As discussed in section 5.1, if all new licences are released as unrestricted taxis, an additional 280 taxis on the road on Friday and Saturday nights would give a 10% reduction in licence lease costs. However, the release of PALs only would allow a lot more taxis on the road during this time – for the same impact on the lease income from unrestricted perpetual licences.

In addition, releasing more PALs compared to unrestricted licences will address concerns about further reducing occupancy during off-peak day shifts because PALs are not allowed on the road before midday. If all new licences are released as unrestricted taxis, occupancy during day shifts would fall.

Figure 5.2 shows the improvements in waiting times for different release scenarios. If PALs are included for release, there will be more taxis on the road to meet the Friday and Saturday peak for a given impact on licence values.

110 Transport for NSW website, Annual taxi licence release tenders,

http://www.transport.nsw.gov.au/taxitender. 111 Taverner Research, Survey of Taxi Use in Sydney, November 2012.

http://www.transport.nsw.gov.au/taxitender



Figure 5.2 Waiting times under different release scenarios


5.2.3 Releasing more PALs could improve the alignment between supply of taxis and demand for taxis

The supply of and demand for taxis are not currently well aligned. High fixed costs of providing taxi services mean that the supply of taxis on the road is relatively fixed despite demand varying markedly during the week, as the taxis are incurring costs such as licence lease costs and insurance whether they are on the road or not, so operators see it as better to have them on the road as much as possible.

Based on the CIE’s survey results, Table 5.2 shows that the fixed costs of providing taxi services for unrestricted taxis in the 4th quarter of 2011 were $52,188, including $28,789 plate lease costs (ex GST).112 At our public hearing, several stakeholders said that this meant that taxis need to be on the road for most of the week:

We want it on the road as much as possible. In fact an operator of a taxi…. has to have it on the road financially to make a business of it for up to 14 shifts, and that is seven days113

You try to get every cab on the road every day to make money. It was just too hard - if you miss two or three shifts, without driving it yourself, you can't afford to keep it. You can't afford to run it.114

112 In 2011/12 dollars. Fifty per cent of maintenance and insurance costs have been allocated as

variable costs. In the taxi industry model, taxi costs have been updated to 2012/13 dollars by inflating them by the Taxi Cost Index used in the 2012 fare review.

113 Comments made by M. Burrage, Transcript - Review of Sydney Annual Taxi Licence release 2013/14 - 24 October 2012, p 24, http://www.ipart.nsw.gov.au/files/c04d5acf-5cd8-4cd2-a848-a0f600be6206/Transcript_-_Review_of_Sydney_Annual_Taxi_Licence_release_2013-14_-_24_October_2012.pdf.

4

5

6

7

8

9

10

11

Mon-Fri Day Mon-Thurs Night Sat-Sun Day Fri-Sat Night Sun Night

min

utes

Current 280 unrestricted 160 unrestricted, 200 PALs 480 PALs Open entry (+8,000)

http://www.ipart.nsw.gov.au/files/c04d5acf-5cd8-4cd2-a848-a0f600be6206/Transcript_-_Review_of_Sydney_Annual_Taxi_Licence_release_2013-14_-_24_October_2012.pdf







This is consistent with the CIE’s survey results, which show that most taxis operate an average of between 10 and 11 shifts out of the 14 possible shifts in a week.115

Having some single-shifted taxis would allow the supply of taxis to better meet the demand throughout the day – but this is only viable with substantially lower fixed costs. The fixed licence lease costs for unrestricted licences will not reduce enough in the short term because of the need to avoid unreasonable impacts on existing licence owners. However, we estimate that a 25% reduction in licence lease costs over a 5-year period should be enough to allow unrestricted taxis to reduce the number of shifts they need to operate just to remain viable to around 8 shifts per week.

In the interim, releasing PALs should allow more taxis to operate a viable business as a single-shifted cab. As Table 5.2 shows, this is because the fixed costs for PALs are around 20% lower than for unrestricted licences, including a lower licence cost reflecting their reduced opportunity to generate revenue (in practice they operate up to 7 shifts instead of 14).

This provides a better match of supply and demand while we transition to lower unrestricted licence values. It also gives drivers and operators who wish to run their own taxi business without arranging other people to drive for them the opportunity to enter the market.

The existing operating patterns of PALs suggest they are currently viable for 6-7 10-hour shifts. A reduction in the annual licence costs for PALs should mean that they become viable to operate fewer than 6 shifts a week.

114 Comments made by G. Seisun, Transcript - Review of Sydney Annual Taxi Licence release 2013/14 -

24 October 2012, p 26. 115 The CIE, Reweighting of the taxi cost index – final report, April 2012, p 31.





Table 5.2 Fixed costs of providing taxi services (ex GST, in $2011/12)

Fixed costs Unrestricted licence

PAL Differences between unrestricted licences and PALs

Maintenance costs (50% fixed) $3,781 $3,781 Insurance (50% fixed) $6,582 $6,582 Vehicle lease payments $5,805 $2,902 50% - Vehicle retains greater

value or has longer life when driven less. Fixed costs may be spread over taxi use and non-taxi use (may be driven as family vehicles).

Network fees $7,231 $7,231 Current plate lease costs $28,789 $21,600 Around 75% of the costa

Total $52,188 $42,095 a Based on the relativities between the median PAL annual licence fees and the median unrestricted annual licence fees. Transport for NSW website, http://www.transport.nsw.gov.au/taxitender. Source: The CIE, Reweighting of the taxi cost index – final report, April 2012, p 9.

5.2.4 Comments from submissions on the type of licences to be released

The 2 main comments made on the release of Peak Availability Licences were: they won’t work as a different business model because their fixed costs are the

same as an unrestricted licence116 they won’t address the peak problem because they can be driven at non-peaks

(Sunday and Monday nights) as well as peaks (Friday and Saturday nights).117

The NSW Taxi Council submitted that the recommendation on PALs contravened the term of reference that we should have regard to ‘simplifying existing taxi licence structures’.118

Some submissions suggested that any new PALs to be released should have different hours of operation from the current PALs: Friday and Saturday evening work only was the most common suggestion.119

116 M Burrage submission, 12 January 2013; Legion Cabs submission, 15 January 2013. 117 Manly Warringah cabs submission, 9 January 2013; NSWTDA submission, 21 January 2013, p 3. 118 NSW Taxi Council submission, 21 January 2013, pp 17-18. 119 M Gordon submission, 18 January 2013; Anonymous submission (W13/67), 18 January 2013; D

Lipski submission, 21 January 2013, p 24.




Former Victorian Taxi Inquiry Commissioner David Cousins observed that the Victorian Taxi Inquiry had taken the view that the taxi market should have maximum flexibility to respond to market demands and that peak/off-peak supply issues were better addressed through differential fares rather than licence hours of operation. He also suggested that having licences with different conditions created regulatory enforcement issues.120 Manly Warringah Cabs’ submission suggested that PAL cabs should be identified by decals or different coloured plates to assist regulatory enforcement.121

Manly Warringah Cabs also suggested that some of the new licences should have a requirement to attach to regional networks (ie, Manly Warringah or St George) to improve taxi coverage in these areas.122

5.2.5 Response to comments on the type of licences to be released

We are aware that PALs are not just available for driving at peak times, but also at off-peak times. We also agree that in theory the market is best left the flexibility to decide when and where cabs should be on the road (so we do not support the Manly Warringah Cabs proposal). However, in practice, this particular market is not very flexible and traditional practice and high fixed costs lead participants to try to get every cab on the road as much as possible.

The main advantage of a PAL for encouraging a different approach to be taken is that it can only be single-shifted, and it has to operate under a different business model which requires lower fixed costs. We encourage would-be operators to carefully assess the cost structure of a PAL taxi business and bid an appropriate amount for tendered PALs. We note that the NSW Government has a range of assistance available to people wanting to start a small business, including potential taxi operators.123

We appreciate that submissions in favour of peak licences that can be driven on Friday and Saturday nights only are trying to focus a solution on the times of the week that have the biggest shortages of cabs (Friday and Saturday nights). However, as discussed above, we favour as few restrictions as possible to allow the market to address shortages, and see PALs as a transition measure. A new peak licence would also complicate the existing licence structure. We do not consider that issuing more of an existing licence type undermines the objective of simplifying existing licence structures.

We sought advice from TfNSW on the identification of additional PALs. TfNSW advised that PALs are already identified by their number plates, which are in the form ‘T 9xxx’.

120 D Cousins submission, 15 January 2013, p 3; NSW Taxi Council submission, 21 January 2013. 121 Manly Warringah Cabs submission, 9 January 2013. 122 Manly Warringah Cabs submission, 9 January 2013. 123 See www.smallbiz.nsw.gov.au for more information.

http://www.smallbiz.nsw.gov.au/



5.3 Estimated effects of releasing 160 unrestricted licences and 200 PALs in 2013/14

We used the taxi industry model developed by the CIE to estimate the effects of releasing 160 unrestricted licences and 200 PALs124 in 2013/14 on the performance of taxi services, on operators and drivers, and on existing holders of perpetual licences.

We estimate that:

the performance of taxi services, as experienced by passengers, will improve, with lower waiting times, especially at peak times of the week (Friday and Saturday nights), and more taxi trips taken

operators will have access to cheaper licences, and in particular will have access to a more affordable and simpler business model through the PALs

with more licences, the demand for taxi drivers will be higher and they should be able to earn more per hour.

Our estimates of these effects are discussed in more detail below.

These estimated outcomes are based on our expectations of what will occur if Transport for NSW releases the number of licences we recommend, assuming fares are held constant in nominal terms and other costs of providing taxi services increase by the rate of inflation.

As set out in more detail in Appendix C, the taxi industry model estimates the outcomes of a change to licence numbers in the long run, and does not show the transition path to the outcomes. Nor can the model tell us how long it might take to get there. A number of submissions raised concerns over transition effects, which we discuss in more detail below.


225 unrestricted licences and 230 PALs.



5.3.1 Estimated effects on performance of taxi services

With the release of 160 unrestricted licenses and 200 PALs,125 waiting times for taxis should be lower, with an estimated improvement of 3% to the average waiting time across all shifts and 6% improvement in average waiting times on Friday and Saturday nights. We estimate that there will be a 6% increase in the number of total trips taken.

Some of the additional taxi trips will be due to growth in population, household or business income, economic activity or tourism activity, but some will be due to people using taxis who otherwise would not have because waiting times are lower.

Comments from submissions on effects on performance of taxi services

Many submissions expressed the view that more licences at a lower price will attract drivers who will not provide a high level of service to the public, or operators will cut corners on maintenance to save money, or more competition for the same work will induce drivers to behave badly to chase fares.126

On the other hand, the ATDA’s submission on our issues paper suggested that a price that allows an operator-driver model will ‘substantially improve the safety and security of taxi drivers and taxi passengers with a new resource of professional drivers.’127

The NSW Taxi Council stated that double-shifting of taxis and pressure on operators to have a taxi on road as much as possible (including forcing drivers to take off-peak shifts bundled with peaks) is good for passengers and off-peak coverage, so changing to a single-shifted norm will make off-peak service worse.128

The NSW Taxi Council and others also submitted that waiting times are good already, and that a less than 1 minute improvement in waiting times does not seem to be a big enough improvement to justify disrupting the industry.129


225 unrestricted licences and 230 PALs. 126 Anonymous submission (W12/1429), 12 December 2012; Anonymous submission (W13/3),

3 January 2013; M Kiani submission, 8 January 2013; M Gordon submission, 18 January 2013; G Pavlis submission, 20 January 2013; P Miller submission, 20 January 2013; D Lipski submission, 21 January 2013, p 24; Anonymous submission (W13/90), 21 January 2013; N Issa submission, 22 January 2013.

127 ATDA submission, 2 November 2012, p 1. 128 NSW Taxi Council submission, 21 January 2013, p 28. 129 NSW Taxi Council submission, 21 January 2013, p 17; S Porcaro submission, 21 January 2013,

p 2; D Lipski submission, 21 January 2013, pp 5-6; A Yazdabadi submission, 21 January 2013.



Response to comments from submissions on effects on performance of taxi services

We do not accept that a high licence lease price is an effective form of quality regulation, or that cheaper licences will attract lower-quality drivers or operators.

We also consider that the industry can support and would benefit from a mix of business models, rather than being all double-shifted or all single-shifted, in order to better match supply with demand.

Nor do we accept that a 3% improvement in average waiting times, and a 6% improvement in average waiting times for passengers on a Friday and Saturday night is ‘not much’, or that the number and mix of licences we are recommending will disrupt the industry.

5.3.2 Estimated impact on operators and drivers

Entry into the taxi market should be more affordable for operators, who should have an estimated 10% lower annual licence costs. There should be the opportunity to run a different business model with a single-shifted cab with a PAL and no need to hire other drivers.

We also consider that drivers who do not operate their own taxi should be better off. With more licences, the demand for taxi drivers will be higher and they should be able to earn more per hour (operators will have to discount pay-ins to get drivers, so drivers’ costs would go down even if their revenue doesn’t increase.) In addition, releasing more PALs compared to unrestricted licences will address concerns about further reducing occupancy during off-peak day shifts because PALs are not allowed on the road before midday.

Comments from submissions on impact on drivers and operators

Many submissions expressed concern that drivers would be worse off because more cabs will be competing for the same amount of work.130 Even where submissions accepted that more licences should increase demand through lower waiting times, and external sources of demand should also be growing, they expressed concern that the flow-through of costs to licence owners would be too

130 P Fletcher submission, 12 December 2012; Anonymous submission (W13/3), 3 January 2013; P

Andersen submission, 13 January 2013; T Fathinia submission, 14 January 2013; Legion Cabs submission, 15 January 2013; J Barber submission, 16 January 2013; Anonymous submission (W13/56), 17 January 2013; M Gordon submission, 18 January 2013; E Mollenhauer submission, 21 January 2013; NSWTDA submission, 21 January 2013, p 1; H Batth submission, 21 January 2013; Anonymous submission (W13/67), 18 January 2013; P Miller submission, 20 January 2013; S Porcaro submission, 21 January 2013; D Lipski submission, 21 January 2013, p 23; St George Cabs submission, 21 January 2013; A Yazdabadi submission, 21 January 2013; Anonymous submission (W13/90), 21 January 2013; N Issa submission, 22 January 2013; ATDA submission, 23 January 2013, pp 5-6.



slow and that the transition effects of change would be too damaging to drivers.131

Submissions were also concerned about the impact on operators, particularly on operators’ ability to get drivers.132 Others expressed scepticism that lower annual licence costs would mean lower costs for operators, predicting that other costs (such as network fees) would go up to compensate.133

Response to comments from submissions on impact on drivers and operators

Most of the concerns expressed in submissions relate to transition effects. In the short term, a change to fares or a change to the number of cabs on the road could have an impact on driver income. However, in the longer term, driver incomes are determined by supply of and demand for driver labour, rather than by the level of fares or revenue from fares, and with more licences, demand for driver labour will be higher.

Operators also might experience short-term impacts when they have to discount pay-ins to attract drivers. However, the improved access to cheaper licences, through the opportunity to bid for annual licences and any market reaction to that, would off-set the lower pay-ins. The timing will be dependent on how quickly licence lease costs reduce.

5.3.3 Estimated impacts on existing holders of perpetual licences

Our task is not to determine the impact that licence owners will experience, but to recommend a number of licences to be released without having an unreasonable impact on existing licence holders. The impact is an effect of releasing licences, and we are required to balance this estimated effect with the other objectives of the reforms.

In our issues paper we asked for feedback on what constitutes an unreasonable impact on licence holders.

131 NSW Taxi Council submission, 21 January 2013, p 25, ATDA submission, 23 January 2013, p 6. 132 M Mikhail submission, 29 December 2012; Anonymous submission (W13/3), 3 January 2013; S

Guy submission, 16 January 2013; J Barber submission; 16 January 2013; NSWTDA submission, 21 January 2013, p 3; RSL Cabs submission, 18 January 2013; St George Cabs submission, 21 January 2013.

133 P Miller submission, 20 January 2013; A Yazdabadi submission, 21 January 2013; N Issa submission, 22 January 2013.



Many submissions addressed this issue. Some stakeholders felt that any investment comes with risk.134 One said that a taxi licence investment should not be protected by the Government.135 Others felt that any material impact would be unreasonable.136

Current owners argued that their taxi licence was equivalent to superannuation,137 or a house purchase.138 They generally agreed that, in the words of one submission, ‘the equity of the plate is paramount’.139

Some submitters accepted that past returns have reflected sovereign risk (that is, a taxi licence only has value because of Government policy, and any change to Government policy could affect that value.)140

Government policy did change in 2009, with the 2009 reforms providing for additional licences to be released each year with the view to providing more affordable licences.

Despite existing owners’ focus on the current paper value of their licence, we consider that impact should be assessed by looking at the income owners can receive from the licence (ie, lease payments). This is because:

the cost of the licence to the owner varies depending on when it was purchased and is not related to the current market value

the current market value is determined by investors’ expectations of future returns, can be quite volatile and is harder to predict than changes to lease income.

In this section we look at the potential impacts on income from perpetual unrestricted and PALs.

134 NSWTDA submission, 9 November 2012; ATDA submission, 2 November 2012; N O’Brien

submission, 9 November 2012. 135 N O’Brien submission, 9 November 2012. 136 E Atra submission, 5 November 2012; D Padgett submission, 23 October 2012; S Hambly

submission, 21 November 2012. 137 P Miller submission, 24 October 2012; P Zarzavtjian submission, 29 October 2012; H Easson

submission, 9 November 2012; St George Cabs submission, 9 November 2012. 138 M Cohen and D Holland submission, 27 October 2012. 139 P Louridas submission, 5 November 2012. 140 NSW Taxi Council submission, 9 November 2012, p 17; ATDA submission, 2 November 2012.



Income from perpetual unrestricted licences

Owners of perpetual unrestricted licences can earn income from leasing their licences. Our estimate of the current average annual cost of a lease is $28,155 (not including GST, in 2012/13 dollars), based on data from the survey of taxi operators conducted by the CIE in October 2011, deflated by 2.2% for the change in lease values since the end of 2011.141

We estimate that if 160 unrestricted licences and 200 PALs142 are released in from July 2013, income from perpetual licences could fall by 10%.

We note that the Victorian Taxi Inquiry’s final report recommended making unrestricted taxi licences available on demand in Melbourne for $20,000 (not including GST), a reduction of 33% on the market assignment price of $30,000 (not including GST).143

As our task is to set licence numbers rather than lease prices, we have taken a more conservative position than the Victorian Taxi Inquiry in terms of expected impacts. However, we also consider that in order to meet the objectives of the legislation, more licences will need to be released in future years and, as a result, we have considered the reduction in lease costs that would be appropriate over a 5-year period. We have developed a proposal to release 160 new unrestricted licences and 200 new PALs144 in the context of a possible reduction in licence lease costs of around 25% over the next 5 years.

Income from leasing perpetual PAL licences

About half the existing 286 PALs (as at 1 January 2013) are perpetual licences. There have been around 20 trades of perpetual PALs in the last 5 years, at around $250,000 - $300,000 (Figure 5.3 – nominal values). These transfer values are around 75% of the transfer values for unrestricted licences. This is consistent with annual PAL fees, which are also around 75% of the fees for unrestricted annual licences.145

141 Change in lease values calculated based on the change in average annual licence price for all

licences on foot as at December 2012 compared to December 2011; information provided by Transport for NSW. The CIE, Reweighting of the taxi cost index – final report, April 2012, p 9.

142 Before the adjustments discussed in Chapter 6. The total numbers recommended for release are 225 unrestricted licences and 230 PALs.

143 Victorian Taxi Inquiry, Customers First: Service, Safety, Choice, May 2012, p 471 and 518, http://www.taxiindustryinquiry.vic.gov.au/__data/assets/pdf_file/0004/67873/09-Part-E-20120626.pdf.

144 Before the adjustments discussed in Chapter 6. The total numbers recommended for release are 225 unrestricted and 230 PALs.

145 Transport for NSW website, Annual taxi licence release tenders, http://www.transport.nsw.gov.au/taxitender.






Figure 5.3 PAL transfer values 2001 – 2012 (nominal)

Data source: Transport for NSW.

We do not have information on the current market lease value of perpetual PALs, but we would expect these lease values to be close to the licence fees for annual PALs. We have estimated that if all the additional licences are released as unrestricted licences, PAL lease costs will fall by around 10%. If a proportion of licences are released as PALs, the lease costs might fall by more, depending on whether operators are able to obtain a licence significantly cheaper in the annual tender. However, we expect that these decreases will affect a very small number of investors.

$0

$50,000

$100,000

$150,000

$200,000

$250,000

$300,000

$350,000



Comments from submissions on impact on licence owners

We received many submissions from licence owners (some of whom were also operator/drivers, or were retired operator/drivers), objecting to our recommendations because of the impact on their licence value and/or lease income. All said that 10% decrease in lease income was too much.146 Three said that any deliberate reduction in licence income is an unreasonable impact on a licence holder.147

Contrary to our description of little change to the cost of entry to the industry since the 2009 reforms, several argued that owners are already worse off as a result of the 2009 reforms, with lease rates going down and licence values at lower levels at the end of 2012 compared to the end of 2011.

Some owners argued that we should regulate other input costs (such as registration and insurance) to make operating a taxi more affordable.148

On the other hand, two submissions said that our recommendations did not go far enough in reducing rents to licence owners,149 pointing out that the Victorian Taxi Inquiry is recommending an immediate one third drop in lease prices.

Response to comments from submissions on impact on licence owners

No arguments were advanced from owners as to what an acceptable alternative impact would be (apart from ‘none at all’.) The government made the decision in 2009 when the legislative amendments were made that changes would be made to the industry that would have an impact on owners, and we consider that an estimated 10% impact on lease income is not unreasonable in the context of the other estimated positive outcomes for passengers, drivers and operators.

146 P Fletcher submission, 12 December 2012; P Andersen submission, 13 January 2013;

Anonymous submission (W13/3), 3 January 2013; Anonymous (W13/34), 13 January 2013; J Barber submission, 16 January 2013; Anonymous submission (W13/56), 17 January 2013; J Klessaut/E Klessaut/T Svircevic submissions, 17 January 2013; M Gordon submission, 18 January 2013; D Petrovic submission, 21 January 2013; Anonymous submission (W13/69), 19 January 2013; P Miller submission, 20 January 2013; S Porcaro submission, 21 January 2013; E Ieraci submission, 21 January 2013; R Palmer submission, 21 January 2013; A Yazdabadi submission, 21 January 2013; P Nicolopoulos submission, 21 January 2013; N Issa submission, 22 January 2013.

147 J Barber submission, 16 January 2013; M Gordon submission, 18 January 2013; P Nicolopoulos submission, 21 January 2013.

148 Legion Cabs submission, 15 January 2013; J Barber submission, 16 January 2013; G Pavlis submission, 20 January 2013; S Porcaro submission, 21 January 2013; H Batth submission, 21 January 2013.

149 ATDA submission, 23 January 2013, pp 7-8; N O’Brien submission, 21 January 2013.



5.3.4 Impact on holders of current annual licences

We also considered impacts on holders of current annual licences who, in the 3 years to date, have bid the following average amounts for a licence. (Tenderers are not charged GST for licences):

Table 5.3 Average successful unrestricted licence bids (nominal $)

2010 2011 2012

Average successful unrestricted licence bid 33,719 33,462 Not available Note: Includes both driver-only and open tenders. Source: Transport for NSW website, < http://www.transport.nsw.gov.au/taxitender> 7 December 2012.

These amounts, higher than the market rate for leases, appear to be based on expectations that lease and licence values will continue to rise. While these are annual licences, they are renewable by right at the same price as bid each year for up to 9 renewals (10 years total term). In an environment where lease values and licence prices are falling, a price fixed in an environment of rising prices becomes too high.

A price adjustment mechanism does exist: annual licence-holders can always bid for a new licence in the annual tenders, and if they obtain a new one for a lower price, can then not renew the original one. This ‘churn’ is a means for the market to be more responsive to information about prices. We discuss the churn mechanism further in Chapter 6.

5.3.5 Transition effects and the viability and sustainability of the taxi industry

As we noted above, the taxi industry model estimates long-run outcomes and does not show what happens while the industry is in transition to equilibrium, nor how long this transition might take.

Comments on transition effects and the viability and sustainability of the taxi industry

A number of submissions were concerned about transition effects and their impact on the taxi industry as a whole.150

The NSW Taxi Council described a process where drivers would not be able to make enough money and would leave the industry, where operators would go out of business because they can’t find drivers, or cut corners and behave dangerously or illegally to stay in business. The NSW Taxi Council asserts that the only way private lease costs can fall substantially is for taxi operators to go out of business and hand back their leased licences.151

150 NSW Taxi Council submission, 21 January 2013, p 25, ATDA submission, 23 January 2013, p 6. 151 NSW Taxi Council submission, 21 January 2013, p 25.




Response to comments on transition effects and the viability and sustainability of the taxi industry

While the NSW Taxi Council’s scenario is possible, we do not consider that it is inevitable. The transition process largely depends on the timing of the adjustment of lease costs. The NSW Taxi Council assumes that lease costs will adjust only slowly; we consider that they could adjust more quickly. We consider that the NSW Taxi Council could also help the transition by working with networks (who manage many licences on behalf of owners) and owners to transition to lower market rates for leased licences without undue disruptions to drivers and operators.

History has shown that as fares have risen steadily each year, profits have passed through to licence owners in the form of higher lease payments, so we consider that changes to the industry that lower revenue should also pass through to licence owners as lower lease payments.

Issuing more licences as PALs should also help to accelerate the transition to lower cost licences, by offering an alternative more affordable business model immediately.

There will be the opportunity to review progress when the determination of annual taxi licences for 2014/15 commences. However, given that new licences for the current determination will only commence from the second half of 2013, only a preliminary assessment of any changes to the industry will be possible.

Nevertheless, it is important to emphasise that transition effects are by definition temporary and that the change is in the long term interests of the industry as a whole. As former Victorian Taxi Inquiry Commissioner David Cousins pointed out in his submission:

a clear distinction needs to be made between the viability and sustainability of individual industry participants and the industry as a whole. The industry may be quite viable and sustainable whilst some members of it are unable to operate profitably. In competitive industries, entry and exit are a normal and healthy aspect of market operation.152

5.3.6 Impact on fares

One of the objectives in our terms of reference is that licence numbers should place downward pressure on fares over time. We are looking more closely at fare issues now we have commenced our review of fares to apply from July 2013.

152 D Cousins submission, 15 January 2013, p 3.



Comments in submissions on impact on fares

Several submissions suggested that other input costs (such as vehicle taxes and registration, green slips and tolls) should be regulated rather than lease prices reduced in order to allow fares to fall.153 Two others suggested that fares should be reduced first to increase demand, and only then should we consider releasing more licences if demand had grown.154

The NSW Taxi Council commented that fares are exactly where IPART has set them over the years, and that we should not have allowed supply of licences to affect our fare recommendations. The NSW Taxi Council submits that if we go back to using actual lease cost changes as the inflator in the Taxi cost indices, there will be downward pressure on fares without having to release more licences.155

Response to comments in submissions on impact on fares

We do not have any responsibility for regulating other input costs, such as registration or motor vehicle taxes, nor is it appropriate for them to be regulated differently for the taxi industry. They are genuine economic costs of doing business rather than an economic rent.

We are considering the impact of the changes to licence numbers during our review of fares. We will also consider the impact that changing fares is likely to have on passengers, drivers, operators and licence owners. We have tested various different fare change scenarios, as set out in Figure 5.4 below. For example, if fares were to increase in line with inflation in 2013/14 instead of remaining constant in nominal terms, we estimate that:

there would be half as many additional total trips compared to keeping fares constant in nominal terms (a 3% rather than a 6% increase) (this is because when fares are more expensive, fewer people catch taxis)

waiting times would be shorter on Friday and Saturday nights (8% better on average compared to 6%) (this is because when fares are more expensive, fewer people catch taxis, so the ones who continue to catch taxis would have more available)

average occupancy per taxi would decrease slightly (by 1%) rather than increase slightly (by 1%) (this is because when fares are more expensive, fewer people catch taxis, so there is less business for each taxi)

153 Legion Cabs submission, 15 January 2013, p 4; J Barber submission, 16 January 2013; G Pavlis

submission, 20 January 2013; S Porcaro submission, 21 January 2013, p 5; H Batth submission, 21 January 2013.

154 M Gordon submission, 18 January 2013; Anonymous submission (W13/67), 18 January 2013. 155 NSW Taxi Council submission, 21 January 2013, p 15.



annual lease income for owners of perpetual unrestricted licences would fall by 7% instead of 10%, and similarly annual lease costs for operators would fall by 7% instead of 10% (this is because the total revenue per taxi is higher if fares are higher, even if occupancy rates are lower, and this is assumed to flow through to the lease cost).

This testing shows the difference in estimated outcomes depending on what happens to fares. Further discussion of these options is contained in our fares issues paper, available on our website.

Figure 5.4 Impacts of our licence recommendation with different fare changes

Data source: Taxi Industry model.

5.4 Could the outcomes be different from our estimates?

All models of markets are estimations of how things actually work. The forecasts produced by a model might not match the real world experience: for instance, there might be changes in elements of the taxi industry model that we have assumed will stay the same, or factors external to the model that affect the outcomes too.

The CIE based the taxi industry model as far as possible on robust sources of data, but for some inputs they have had to use assumptions. We tested the model to assess its sensitivity to some of the key input assumptions: demand growth (which we estimated as 2.5% per annum based on our

assessment of relevant proxies – see Chapter 4 for details)

price elasticity (for which the CIE used a figure of -0.8, based on a review of the literature about price elasticity for taxi services)



value of time (which the CIE estimated as $30 an hour)

change in the costs (other than licence lease costs) of providing taxi services. The CIE based these costs on their survey data from October 2011, updated by the change in the urban Taxi Cost Index for the year to April 2012.

The outcomes of the sensitivity testing are set out below. We found the greatest sensitivity is to the elasticity assumption (largely because the range of estimates is so high). However, after considering the sensitivity testing, we are confident that our modelling outcomes are a reasonable estimate of likely real-world outcomes.

As noted earlier, the taxi industry model cannot predict how long it will take to transition to the forecast outcomes. Furthermore, the change in licence numbers as a result of our recommendations will only be implemented gradually over the second half of 2013.

As a result, only a preliminary assessment of the results of the changes will be possible at the next annual review, which will likely commence before all of the additional licences are on the road. The next annual review will have to consider how to take the preliminary results into account.

5.4.1 Sensitivity to assumption about price elasticity

Figure 5.5 shows that our results are very sensitive to the price elasticity estimate. The CIE assumed a price elasticity to changing real (that is, adjusted for inflation) prices of -0.8 in our taxi industry model, which falls within what they considered (based on the literature) a plausible range of -0.2 to -1.

This means if we hold fares constant in nominal terms, which is equivalent to a real decrease in fares of 2.5%, based on our estimate of the rate of inflation, we expect demand to increase by 2% as a response to cheaper real fares, additional to the underlying growth from external factors.

The Victorian Taxi Inquiry used a price elasticity estimate of -1, which was derived from elasticity modelling for different customer segments undertaken by David Hensher based on stated preference surveys in Melbourne. If we assume an elasticity of -1, we estimate that licence lease costs would drop by around 8%, and improvements in occupancy and passenger waiting times would be more modest.

However, with an elasticity of -0.2, licence lease costs fall by around 16%, and vehicle occupancy falls instead of rises.



Figure 5.5 Impact of price elasticity on taxi industry outcomes


5.4.2 Sensitivity to assumption about value of time

The CIE advised that the value of time generally used for cost benefit analyses lies between $12 and $18, but they considered that the value of time for taxi users is likely to be higher. Therefore the taxi industry model uses an estimate of $30.156 We note that the Victorian Taxi Inquiry used $60.157 However, the model results are not particularly sensitive to the value of time assumption, as shown in Figure 5.6 below.

Figure 5.6 Impact of value of time assumption on taxi industry outcomes


156 Analysis undertaken by the CIE and reflected in the taxi industry model that can be found on

IPART’s website. 157 Information received from the Victorian Taxi Inquiry.



5.4.3 Sensitivity to assumption about changes in costs of providing taxi services

Our modelling for this review assumes that costs will increase in line with inflation (that is, that real costs will not change). We note that the Transport Workers Union has a hearing with the Industrial Relations Commission in June 2013 seeking to increase driver earnings (around 42% of the costs of providing taxi services) by around 18%.158 If it is successful, the uplift in driver costs would increase total costs by around 7.5%. Figure 5.7 shows that if costs increased by 10%, then licence lease costs would fall by over 30%.

Figure 5.7 Impact of changes in costs on taxi industry outcomes


158 ATDA submission, 23 January 2013, p 7 and attachment.

6 Making an allowance for replacement licences and Wheelchair Accessible Taxi licences



Some annual licences that were issued in previous years have been handed back to TfNSW or not renewed. In order to maintain the existing number of licences an adjustment needs to be made to account for the licences that will be handed back or not renewed next year. In addition, as Wheelchair Accessible Taxi licences are available on demand from Transport for NSW and they compete for some of the same work as unrestricted licences, we need to release fewer annual taxi licences through the tender process to account for the WAT licences that will be released. This chapter considers how many additional unrestricted licences and PALs should be offered for tender in order to take these issues into account.

As we expect licence prices to fall over time, the rate at which annual licences are not renewed is likely to increase as more people make use of the annual tender process to try to get a better price for a licence. This chapter also considers what arrangements TfNSW should put in place to assist operators to reduce their costs using this mechanism.

6.1 Additional licences that need to be released

We are recommending adjustments to the number of licences offered for tender in order to account for:

Annual licences that we expect will no longer be operated – specifically, we recommend that Transport for NSW offers additional licences (both unrestricted and PALs) to replace previously acquired annual licences that will be handed back or not renewed (including new licences effectively swapped for existing licences to get a better price).

The expected issue of additional Wheelchair Accessible Taxi licences (WATs) outside the annual tender process – specifically, we recommend that fewer licences need to be issued through the tender process as a result of additional WATs being issued.

A Fringe Area Licence that has not been renewed – specifically, we recommend that an additional Fringe Area Licence be offered for tender as a direct replacement for this licence.



Recommendations

4 Transport for NSW should offer for tender an additional 65 unrestricted licences and 30 peak availability licences in 2013/14 to replace licences we expect to be handed back or not renewed. This includes an adjustment to account for the expected issue of additional Wheelchair Accessible Taxi licences outside the annual tender process.

5 Transport for NSW should offer 1 Fringe Area (Richmond/Windsor) Licence for tender in 2013/14.

6.1.1 Annual licences that we expect will no longer be operated

Every year some annual licences are handed back or not renewed. There are a number of reasons that a licence holder might hand back a licence, including that they have been successful in obtaining a cheaper licence in a subsequent tender release. We consider that the ability to easily relinquish a licence in favour of a cheaper option is an important feature of the current arrangements.

The Act requires the annual licence renewal price for subsequent years to be equal to the price in the first year, with no scope for a reduction in licence fees to be built in. This means that if annual licence costs are falling rather than rising, as we expect to happen in future years, bidding a lower value in a future tender and handing back the more expensive licence is the only means by which operators can reduce the costs of holding an annual licence from TfNSW.

Licences handed back are no longer in circulation and as a result, some adjustment needs to be made to the number of licences released to ensure that licences lost through attrition do not reduce the number of available licences.

We consider that an adjustment for attrition should be made on a prospective basis. We consider that an additional 100 licences (70 unrestricted and 30 PALs) is a reasonable estimate of the number of these licences that will cease to operate during 2013. We formed this estimate based on the number of licences handed back during 2012 (71 unrestricted licences and 31 PALs159).

6.1.2 Expected release of Wheelchair Accessible Taxi licences

In line with the 2009 amendments, Transport for NSW issues Wheelchair Accessible Taxi (WAT) licences on application for $1,000 per year. PwC’s approach made an adjustment to the modelled number of growth licences to account for WAT licences released during the year, as these are another source of taxis that can substitute for taxis with standard licences. PwC deducted 0.45 of a licence for each additional WAT, on the assumption that 90% of WAT jobs are standard jobs that could be performed by any taxi, and that a WAT only operates

159 Information provided by Transport for NSW.



50% of the time (because WAT licence conditions mean they are usually single-shifted rather than double–shifted).160

Stakeholders that raised this issue generally felt that WAT licences are equivalent to more than 45% of an unrestricted licence and so that a greater adjustment should be made to account for them than has been made in the past. For example, the NSW Taxi Council submitted that because WATs tend to operate in peak periods the 45% adjustment is too low.161 Other submissions supported the assumption that a high proportion of WAT jobs are not wheelchair jobs.162

While we consider that arguments can be made regarding the formula to be used, there is little information that would allow a more precise calculation and therefore, a consistent approach to adjusting for new WATs is appropriate. The number of WATs released during 2012 is small, and as a result, changes to the assumptions used in this calculation do not make a significant difference to the number of licences recommended.

A net 12 WAT licences were added to the Sydney taxi fleet in 2012. Assuming WATs are equivalent to 45% of an unrestricted licences we recommend subtracting 5 licences to account for additional WATs which we expect to be released outside this process.

6.1.3 Non-renewed Fringe Area Licence (Richmond/Windsor area)

Transport for NSW has advised that an annual Fringe Area (Richmond/Windsor) licence was not renewed at its renewal date in September 2012.

6.2 What should be done to help operators relinquish a more expensive licence in favour of a cheaper one

To date, every tender has resulted in annual licences with a higher average (mean) price than leases for perpetual licences available through taxi networks. This suggests that bidders for annual licences (which have a fixed annual renewal price which cannot be indexed) have been factoring in an expectation that lease prices would continue to rise. Once it becomes clear that the annual licence release process will continue to add new licences to the fleet each year, we expect bidding behaviour to change. We also expect that greater numbers of existing annual licence holders will bid in subsequent licence releases and obtain a licence at a lower annual cost than the one they already hold.

160 PwC, Annual taxi licence release 2012/13 – Final Report, 15 March 2012, pp 17-18. 161 NSW Taxi Council submission, 9 November 2012. 162 ATDA submission, 2 November 2012, NSWTDA submission, 9 November 2012, St George Cabs

submission, 9 November 2012.



As noted above, we consider that the ability to obtain a cheaper licence in a subsequent tender and relinquish the more expensive licence without interrupting the operation of the taxi service or requiring the licence holder to pay for 2 licences is vital to the success of the 2009 reforms. This process is the only means by which operators can reduce the costs of holding an annual licence from TfNSW. When annual licence costs are falling, as we expect to happen under our recommendations, the smooth operation of this process is very important.

The tender information packages issued each year by TfNSW currently state that “Where a licence holder obtains a new licence under a subsequent tender process, TfNSW will make every reasonable effort to align expiry and commencement dates to allow for continuity of operation.”163

Comments from submissions on the churn mechanism

Several submissions said that the churn mechanism would be too costly, quoting $4000 for operators to churn because of insurance being cancelled/reissued, new decals, radio changes and mismatched licence periods.164 Two submissions said that the pool of new licences, even including a churn allowance, was too small to have an impact on prevailing market prices.165

Response to comments from submissions on the churn mechanism

We sought advice from Transport for NSW on churn costs. Transport for NSW estimated a cost imposed by Roads and Maritime Services of around $1,000, providing the operator is using the same vehicle, consisting of the requirement to obtain a blue slip (taxi roadworthiness certificate), and to deregister and register a vehicle.166 These procedures appear to us to be unnecessary red tape and could be eliminated.

In accordance with the tender documentation, Transport for NSW would endeavour to line up the issue of the new licence with the expiry of the previous licence so no additional payments would be payable (noting that successful bidders for licences generally have 6 months to register the licence).

We recommend that TfNSW provides tender documentation that links this more strongly with the opportunity to use subsequent licence tenders as a means of lowering the annual licence cost to an operator.

163 See for example, 2012/2013 Annual Taxi Licence Tender No. TfNSW 2012/018 (Driver Only) Tender

information and conditions, June 2012, p 7. 164 E Mollenhauer submission, 21 January 2013, p 1; T Bradley submission, 22 January 2013; ATDA

submission, 23 January 2013, p 6. 165 E Mollenhauer submission, 21 January 2013, p 1; ATDA submission, 23 January 2013, p 6. 166 Information provided by Transport for NSW.



Recommendation

6 Transport for NSW should ensure that its administration processes and tender documentation support annual licence churn as a positive mechanism operators can use to reduce their annual licence costs.

Recommendation

7 Transport for NSW should remove any unnecessary administrative costs of churn, such as the requirement for a blue slip and deregistration/registration of the same vehicle with a new licence.

7 Process for setting a price and restrictions on who may hold licences



Our terms of reference ask us to make recommendations on the process for setting a price for new licences issued (that is, the auction/tender process) and on whether there should be restrictions on who may bid for licences or on how many licences may be issued to any one bidder.

7.1 Tender process

The legislation offers the choice between an open auction or a sealed tender release; within a sealed tender, the price may be set either by ‘pay as bid’ or ‘lowest successful bid’. To date taxi licences have been issued under a sealed tender and priced using the ‘pay as bid’ method.

Our final decision is the same as our draft decision: that TfNSW should continue with the current tender process. While there are potentially benefits from moving to an approach based on the lowest successful bid (including that this would better accommodate a reserve price), we consider that these are outweighed by the benefits of continuity of approach.

We have also considered other tender process issues, including timing, and the release of market information about tender prices, and recommend that Transport for NSW release preliminary information about tender prices as soon as possible to inform the market.

The following sections discuss these issues in more detail.

7.1.1 Pay as bid or pay lowest successful bid

The pay-as-bid method results in a range of prices for licences. Over time this leads to administrative costs, as every licence holder is charged a different price for essentially the same product. Nevertheless, it does have some advantages including that each applicant putting in a bid knows up-front what they will have to pay if successful, and being able to match this with their willingness to pay.



Setting the price for all licences at the lowest successful bid has the advantage of a uniform price for a uniform product (although this might not reflect the range of values that might exist for the bidders), and it should result in lower average prices and protect to some extent bidders who make over-market price bids from lack of information about prevailing market prices.

However, as the licences are annual, someone who discovers in the pay-as-bid system that they have paid over the market price for a licence has the opportunity to bid for a new licence at a lower price the following year and not renew the original licence. No submissions on our issues paper or participants in the public hearing argued against the pay-as-bid method. However, we received one submission on our draft report that proposed a ‘lowest successful bid’ approach in conjunction with an $8000 reserve price.167

While the lowest successful bid has some advantages, we consider that they are not significant enough to justify moving away from the current approach. The current approach is well understood and now has several years of published results that potential bidders can use to inform themselves. For the sake of stability, we recommend that the pay-as-bid method be retained.

Recommendation

8 Transport for NSW should continue to use the existing tender process, that is, a sealed electronic tender, with pay-as-bid prices for successful tenderers.

7.1.2 Other tender process issues

Tenders have to date mostly been held with a June closing date. Successful tenderers are usually notified in July and have 6 months to complete the registration process and get a car on the road. Transport for NSW does not publish the results of a tender until all licences are issued and on the road, which can be more than 7 months from the tender close date depending on circumstances.

In our view, the regular scheduling of tenders in June (with any unissued licences being retendered as soon as possible) helps the industry to plan and helps to facilitate the churn from a higher-priced licence to a lower-priced one without additional costs.

However, we consider that it would be helpful for the market to be better informed about tender prices sooner, and that Transport for NSW should publish preliminary results of tenders as soon as possible. Given the electronic tender process, we consider that they should be able to do this within 2 weeks of tenders closing.

167 T Hirsch submission, 24 January 2013, p 5.



Recommendation

9 Transport for NSW should continue to release licences in accordance with the same timetable followed in previous years.

Recommendation

10 Transport for NSW should publish the provisional results of tenders for annual licences (for example, by publishing the median and mean successful bids) within 2 weeks after tenders have closed.

7.2 The need for a reserve price

Our final recommendation on a reserve price is the same as our draft recommendation: not to set a reserve price for tenders.

The sections below set out our reasons for not setting a reserve, the comments received through submissions and our response to those comments.

7.2.1 Our draft recommendation not to set a reserve

To date there has not been a reserve price set for tenders but we have considered whether it is necessary to introduce one in order to ensure that the cost of new licences does not result in the market price for leases of perpetual licences falling more than we expect it to.

The lowest successful bid for each tender has not been significantly below the average successful bid. Even the lowest successful bid for the under-subscribed tender (driver-only PALs in 2011) was $20,800, compared to a mean of $24,301.168 Speculative bids appear to have been entirely absent. Given the pay-as-bid method results in a spread of annual licence costs and the historically conservative bidding behaviour of the industry we do not see a strong argument for a reserve price.

We also note that a reserve price would be more compatible with a lowest successful bid approach to pricing. A uniform price determined by the lowest successful bid would allow a reserve price to be set at the estimated impact point. While it is not impossible to set a reserve price under a pay as bid method, the value of doing so is less clear and the process more complicated.

We expect that it will be the average price that has the biggest impact on the market price for leases of perpetual licences, rather than the highest or lowest prices for successful tenders. For this reason, we would need to set the reserve price below the estimated impact point with the aim of having the average bid price above that level. This would require IPART to forecast the distribution of 168 Transport for NSW website, results for Tender No: 2011/008

<http://www.transport.nsw.gov.au/taxitender> accessed 7 December 2012.




bids (without knowing what impact on bidding behaviour the announcement of a reserve price will have).

Comments in submissions on a reserve price

Several submissions supported a reserve price, but for different reasons. Two suggested setting a reserve to maintain the current lease price169 while 2 suggested setting a low reserve to signal what we thought the price should be.170

Response to comments in submissions on a reserve price

Either of the proposals suggested (setting a reserve price to maintain lease prices or to signal a lower price) prevents the market from determining the right price for licences. In our issues paper we did canvass using a reserve price as a floor to prevent ‘unreasonable impacts’ on licence owners, but as we discussed in the draft report, this becomes a very complicated exercise with a ‘pay as bid’ method.

Recommendation

11 No reserve price should be set for licences tendered.

7.3 Reserving some licences for drivers and/or restricting the number that can be granted to the same applicant

Our final recommendation on reserving licences for drivers or restricting the number that can be granted to one applicant is the same as our draft recommendation: not to put restrictions on either of these.

The sections below set out our reasons for not recommending restrictions, the comments received through submissions and our response to those comments.

7.3.1 Our draft recommendation not to place restrictions on bidders

When the 2009 licencing reforms were introduced into Parliament, the Government noted that the primary focus of the changes was on building a business, based on delivering services to passengers. The Government wanted to provide opportunities for lessee operators and experienced drivers to take up a new licence and become their own boss.171

To further these aims, TfNSW may decide to target the issue of a certain number of licences at drivers to provide a career path, or may restrict the total number of licences taken up by any one person or entity in the interests of promoting competition. 169 S Porcaro submission, 21 January 2013, p 4; Anonymous submission (W13/67), 18 January 2013. 170 D Cousins submission, 15 January 2013, p 3; T Hirsch submission, 24 January 2013, p 5. 171 Full Day Hansard Transcript (Legislative Assembly, 30 October 2009, Corrected Copy), p 19076.



There was support in submissions on the issues paper for driver-only allocations and/or for drivers to have access to free or low cost licences.172 In relation to restricting the number of licences that anyone can hold, the NSWTDA submitted that ‘the industry is best served by those people actually doing the work’ and as a result no one should be able to hold more than one.173

We expect that most tenders for licences will come from drivers and operators who will use their licences to provide taxi services. This is because:

The annual licences that have been released since 2010 cannot be transferred, therefore they cannot be held for a period and sold for a profit. This means that they will not be attractive to investors hoping to make a capital gain.

Obtaining a licence for the purpose of leasing it out to an operator is allowed but is only financially attractive if a licence can be obtained from TfNSW at a discount to the rest of the market.

Peak availability licences (PALs) provide a low-cost opportunity for drivers to obtain a licence for their own taxi without the need to contract with other drivers.

Historically, most winning tenderers have been drivers in the open tenders as well as the driver-only tenders.

We are recommending that a minimum of 2.5% of the fleet is offered by tender every year for at least 5 years – that is at least 140 additional licences offered into the market each year. As the market accepts that release of positive numbers of licences will continue over the longer term, there is no reason to believe that drivers and operators would miss out on a licence directly from TfNSW if they wanted one, or be induced to bid above the value they place on these licences. As a result, the scope for others to make money by leasing out new annual licences will not exist.

Prices of annual (10-year) licences are set by demand and supply. Over time, as greater numbers of licences are available, we expect the market value of these licences to fall. This will benefit drivers and operators wanting to obtain a licence (or multiple licences) for the purpose of providing taxi services – providing them with lower cost licences. This should ensure that any driver who wants to own their own licence can afford to do so.

172 ATDA submission, 2 November 2012, J Soothhill submission, 29 October 2012, NSWTDA

submission, 9 November 2012, St George Cabs submission, 9 November 2012. 173 NSWTDA submission, 9 November 2012.



The release of large numbers of PALs provides an opportunity for drivers to obtain low cost licences and to own their own licence without the need to keep the taxi on the road for 24 hours a day, 7 days a week. PALs provide an alternative business model for the industry, particularly for drivers who want to own their own licence and drive their own taxi. We expect bid prices for PALs to be at a significant discount to unrestricted licences.

Provided enough licences are available that drivers and operators have ready access to affordable licences directly from TfNSW, we do not consider that there is any need to restrict the number that can be granted to the same applicant or to reserve licences for drivers only.

Comments in submissions on the tender process

Two submissions174 sought licences to be restricted to one per owner/operator, while 2175 supported the ‘no restrictions’ approach.

Response to comments in submissions on the tender process

The proponents of restrictions argue that they are necessary to achieve the objective outlined above: to provide opportunities for lessee operators and experienced drivers to take up a new licence and become their own boss. However, for the reasons we discussed in the draft report, we do not consider it is necessary to impose restrictions in order to facilitate that business model.

Recommendation

12 No restrictions should be placed on who may bid for licences or on the number of licences that any one person may hold.

174 M Burrage submission, 12 January 2013; ATDA submission, 23 January 2013, p 4. 175 Anonymous submission (W13/1), 3 January 2013; S Porcaro submission, 21 January 2013, p 4.


Annual taxi licence release for Sydney 2013/14 IPART

8

Appendices

7 Process for setting a price and restrictions

on who may hold licences


A Terms of Reference





B List of submissions



B.1 List of submissions on issues paper Name Date

D Allatt 16 October 2012 Anonymous (W12/1038) 11 October 2012 Anonymous (W12/1161) 1 November2012 Anonymous (W12/1176) - confidential 4 November 2012 Anonymous (W12/1177) - confidential 4 November 2012 Anonymous (W12/1244) - confidential 8 November 2012 Anonymous (W12/1250) - confidential 9 November 2012 B Atra 9 November 2012 E Atra 5 November 2012 Australian Hotels Association 5 November 2012 Australian Taxi Drivers Association (ATDA) 19 October 2012 Australian Taxi Drivers Association 2 November 2012 City of Sydney 16 November 2012 Council of Social Service of NSW (NCOSS) 23 November 2012 C Dass (Confidential) 2 November 2012 P Doyle 26 October 2012 H Easson 9 November 2012 Z Fan 16 October 2012 P Fletcher 9 November 2012 S Hambly 21 November 2012 D Holland 27 October 2012 D Lipski 9 November 2012 P Louridas 5 November 2012 M Mancini 16 November 2012 P Miller 24 October 2012 NSW Taxi Council 9 November 2012 NSW Taxi Drivers Association (NSWTDA) 9 November 2012 N O’Brien 9 November 2012 D Padgett 23 October 2012 D Padgett 30 October 2012 Physical Disability Council of NSW 19 November 2012 R Prideaux 26 October 2012



Name Date

S Romijn 12 October 2012 J Rowe 23 November 2012 G Saberton (Confidential) 22 October 2012 St George Cabs Co-Operative Ltd 9 November 2012 J Soothill 29 October 2012 T Svircevic 21 October 2012 Sydney Metro Taxi Fleet 13 November 2012 P Tweddell (Confidential) 18 October 2012 H Waraich 16 October 2012 P Zarzavatjian 29 October 2012

B.2 List of submissions on draft report Name Date

P Andersen 13 January 2013 Anonymous (W12/1429) 12 December 2012 Anonymous (W13/1) 3 January 2013 Anonymous (W13/2) 3 January 2013 Anonymous (W13/3) 3 January 2013 Anonymous (W13/30) 11 January 2013 Anonymous (W13/31) 11 January 2013 Anonymous (W13/34) 13 January 2013 Anonymous (W13/35) 14 January 2013 Anonymous (W13/37) - confidential 14 January 2013 Anonymous (W13/39) - confidential 14 January 2013 Anonymous (W13/47) – confidential 16 January 2013 Anonymous (W13/48) 16 January 2013 Anonymous (W13/56) 17 January 2013 Anonymous (W13/57) - confidential 17 January 2013 Anonymous (W13/64) 18 January 2013 Anonymous (W13/67) 18 January 2013 Anonymous (W13/68) – confidential 19 January 2013 Anonymous (W13/69) 19 January 2013 Anonymous (W13/72) - confidential 20 January 2013 Anonymous (W13/90) 21 January 2013 E Atra 6 January 2013 Australian Taxi Drivers’ Association (ATDA) 23 January 2013 J Barber 16 January 2013 C Barut - confidential 17 January 2013 H Batth 21 January 2013 R Bentley 9 January 2013 T Bradley 22 January 2013



Name Date

M Burrage 12 January 2013 D Cousins 15 January 2013 C Dass 14 January 2013 H Easson 16 January 2013 T Fathinia 14 January 2013 P Fletcher 12 December 2012 D Gill 10 January 2013 M Gordon 18 January 2013 S Guy 16 January 2013 T Hirsch 24 January 2013 E Ieraci 21 January 2013 N Issa 22 January 2013 M Kiani 8 January 2013 E Klessaut 17 January 2013 J Klessaut 17 January 2013 G Koutsioukis 17 January 2013 Legion Cabs 15 January 2013 J Lind 17 January 2013 D Lipski 21 January 2013 Manly Warringah Cabs 9 January 2013 M Mikhail 29 December 2012 P Miller 20 January 2013 E Mollenhauer 21 January 2013 NSW Taxi Council 21 January 2013 New South Wales Taxi Drivers Association Inc (NSWTDA) 21 January 2013 P Nicolopoulos 21 January 2013 N O’Brien 21 January 2013 R Palmer 21 January 2013 G Pavlis 20 January 2013 I Pergamalis 16 January 2013 D Petrovic 21 January 2013 S Porcaro 21 January 2013 RSL Cabs 18 January 2013 St George Cabs Co-Operative Ltd 21 January 2013 T Svircevic 17 January 2013 E Tomko - confidential 21 January 2013 T Tsang – confidential 16 January 2013 A Yazdabadi 21 January 2013

C Taxi industry model



We commissioned consultants the CIE to develop a model of the taxi industry.

All models of markets are estimations of how things actually work. Good models are those that are able to take into account the key features of the market and provide quantitative linkages between changes in certain variables (such as the number of taxi licences) affect other variables (such as the waiting time for taxis). However, there are a number of reasons why real-world outcomes might be different from modelled forecasts: even where a model accurately quantifies the relationship between variables, the forecast inputs might be different (eg, population might not grow by as much as predicted), and the outcomes might be affected by factors that have not been accounted for in the model.

The CIE’s work builds on work done by other economists to model taxi markets, and uses data specific to the Sydney taxi market sourced from Transport for NSW, the Bureau of Transport Statistics, a survey of taxi drivers and operators conducted in late 2011, taxi networks and the Taverner survey of taxi passengers conducted in November 2012.

We are confident that the taxi industry model is a good fit to the data and the relationships between variables. We also tested the model to see how sensitive it was to our assumptions (for example, if price elasticity is actually different from the -0.8 we assumed, how different are the forecast outcomes?). We have published the results of the sensitivity testing throughout this final report so stakeholders can see what the range of forecast outcomes is.

The taxi industry model is a long-run equilibrium model, which means it shows what the market will look like once it has fully adjusted to a change (such as increasing the number of licences). It does not show what the impacts will be as the market works through the changes – how quickly passengers will react to lower waiting times, whether occupancy will be lower in the short term – or how long it will take to reach equilibrium.

While the outputs of the taxi industry model are subject to the uncertainty inherent in any modelling exercise, and the model does not show the timing of the outputs, the model is clear on the direction of the impacts: better outcomes for passengers, drivers and operators, and lower (but not unreasonably lower) lease income from perpetual and ordinary licences.



The elements of the taxi industry model

The inputs to the taxi industry model include: Data from the taxi driver and operator survey conducted by the CIE for us in

2011 (cost structure of providing taxi services, revenue per shift, hours per shift, share of trips with booking fee).

Taxi network data on taxis on the road per shift. This data was used to develop a constraint on the percentage of total licences that would be on the road during a shift. For the draft report, the constraint was 80% for any shift, but for the final report the constraint was applied as ranging between 70% for off-peak shifts to 87% for peak shifts, which was a better fit to the data. This minor change to the parameters of the model resulted in some minor changes to the modelling outcomes.

Current fares.

Data on occupancy rate per kilometre (supplied by the Australian Taxi Drivers’ Association based on 5 million kilometres of travel in 12 cabs).

Bureau of Transport Statistics data on average trip length.

Data on waiting times from the taxi use survey conducted by Taverner.

The taxi industry model incorporates: a demand equation, linking demand for taxi services to the price of services

and waiting time a waiting time equation, linking the average waiting time to the number of

taxis on the road and the share taken up by demand

a taxi entry condition, linking the decision of a taxi to drive a shift to the returns available.

Other inputs are assumptions that can be changed. The model was calibrated using plausible values for these assumptions so that the base case matched the current industry structure in terms of number of licences, lease value and taxis on the road per shift. Two important assumptions are:

Base case price elasticity is -0.8; the Victorian Taxi Inquiry’s draft report suggested an elasticity of around -1 for Melbourne, while Booz Allen Hamilton, in a report for IPART in 2003, noted that the majority of international studies reported a demand elasticity of -0.2 to -1.0.

The base case value of time is $30 per hour which is higher than the range normally used in economic appraisals for public transport projects, on the basis that taxi passengers have a higher value of time than public transport passengers. The results of the model are not very sensitive to this assumption.



For a given number of additional licences (which can be added as unrestricted or Peak Availability licences (PALs) or a mix of both), the model solves for the residual profit of an unrestricted licence, which is equivalent to the lease or annual licence value. On a per shift basis, the model’s outputs include number of trips, taxis on the road and average passenger waiting time.

Glossary


Glossary

ABS Australian Bureau of Statistics

Annual taxi licence A taxi licence obtained from TfNSW which is valid for a year, but can be automatically renewed each year for a term of up to 10 years. The holders of these licences can lease but not transfer (sell) them.

ATDA Australian Taxi Drivers Association

CIE, The The Centre for International Economics

CPI Consumer Price Index: a measure of inflation, or the change in the cost of living over time.

Elasticity, price The amount by which demand changes in response to price changes. For example, if the price elasticity of demand for taxi services is -1, a 10% increase in taxi fares will lead to a 10% drop in taxi use. If the elasticity is -0.2, a 10% increase in taxi fares will lead to a 2% drop in taxi use.

Elasticity, waiting time The amount by which demand changes in response to changes to waiting time. For example, if the waiting time elasticity of demand for taxi services is -1, a 10% increase in waiting time will lead to a 10% drop in taxi use. For the taxi industry model, waiting time was converted to a price using a value of time of $30 an hour, and the same price elasticity value applied.

Fringe Area Licence A taxi licence which must be operated predominantly in nominated geographical areas on the outskirts of Sydney. There are 10 Richmond/Windsor FALs, 2 Richmond Windsor Fringe WATs and 2 Helensburgh FALs.

Glossary


GDP Gross domestic product. GDP is a measure of national output.

IPART Independent Pricing and Regulatory Tribunal. Provides taxi fare recommendations to Transport for NSW on an annual basis.

Latent demand Latent demand represents the additional taxi trips that would be taken if waiting time for taxi services fell, or if fares were reduced

Network, taxi Taxi networks provide a radio booking service to the taxi operators who are affiliated with them, as well as security monitoring services for taxi drivers and passengers. Networks must be authorised by Transport for NSW

Network fee Fee paid by the operator of a taxi to affiliate with an authorised taxi network.

Nominal Not adjusted for inflation. For example, if something cost $100 in 1990 and the same thing cost $100 in 2012, its nominal cost has not changed. However, adjusting for 22 years of inflation, its real cost is lower in 2012 than in 1990.

NSWTDA New South Wales Taxi Drivers Association.

Operator, taxi Taxi operators are responsible for the day-to-day management of one or more taxi cabs. Operators may be individuals or corporations, and must be accredited by Transport for NSW. They must also be affiliated with an authorised network, fit out their vehicle with the network’s livery and install the network’s communications equipment. They must insure and maintain their vehicle(s), and hold a taxi licence for each vehicle they operate.

Ordinary taxi licence From 1990, ordinary licences with 50-year terms were issued by the transport department on request at the prevailing price for perpetual licences. These licences came with the right to renew them at the end of their term, so they are effectively perpetual.

Peak Availability Licence (PAL)

A taxi licence that can only be operated between noon and 5 am.

Glossary


Pay-in The amount paid by a taxi driver to an operator for the use of a taxi. Maximum pay-ins for Sydney taxis are determined by the NSW IRC and set out in the Taxi Industry (Contract Drivers) Contract Determination 1984 but discounting below this rate is common.

Perpetual taxi licence A taxi licence that does not expire. It can be transferred (sold) or leased. New perpetual licences have not been issued in Sydney since 1990.

Price elasticity See elasticity, price

PwC PricewaterhouseCoopers

Real Prices or costs that have been adjusted for inflation. So something that cost $100 in 1990 and $100 in 2012 has had a drop in its real price since 1990. If the measure of inflation (usually CPI) has risen by 30% over that time, the real price of the object in 2012 can be expressed as ‘$76.92 in 1990 dollars’ (or the real price of the object in 1990 can be expressed as ‘$130 in 2012 dollars)

Restricted taxi licence A taxi licence issued with operating restrictions. Examples of restrictions include operating times, operating areas and giving preference to transporting wheelchair users.

Short term taxi licence From the late 1990s, some short-term licences with a term of up to 6 years were issued. Their owners can lease them to another person, but cannot transfer them, and they expire at the end of their term. New short-term licences have not been issued in Sydney since 2009.

STGM Sydney Taxi Growth Model. Method used by PwC to estimate the number of new annual Sydney taxi licences required for each of the years 2010/11, 2011/12 and 2012/13.

Taxi Operator See operator, taxi.

TCIs Taxi Cost Indices. Used by IPART to measure the change in taxi industry costs between fare review periods.

Glossary


Transport for NSW (TfNSW) The NSW Government agency that regulates taxis.

Unrestricted taxi licence A taxi licence issued without operating restrictions.

Waiting time elasticity See elasticity, waiting time

Wheelchair Accessible Taxi licence (WAT)

A licence to operate a Wheelchair Accessible Taxi. The operator is required to give preference to transporting wheelchair users. WAT licences are available on request from Transport for NSW for $1000 a year.

Documents

Annual taxi licence release for Sydney 2013/14 › files › sharedassets › ... · Annual taxi licence release for Sydney 2013/14 . Contents . 1 Introduction : 1 ... the likely