Long run economic and land use impacts of major infrastructure projects Final Report Department of Transport July 2012

Confidential Copy

This report has been prepared on behalf of Department of Transport. SGS Economics and Planning and its associated consultants are not liable to any person or entity for any damage or loss that has occurred, or may occur, in relation to that person or entity taking or not taking action in respect of any representation, statement, opinion or advice referred to herein.

SGS Economics and Planning Pty Ltd ACN 007 437 729 www.sgsep.com.au Offices in Brisbane, Canberra, Hobart, Melbourne, Sydney

Confidential Copy

TABLE OF CONTENTS

KEY TERMS 6 SYNTHESIS AND STRATEGIC IMPLICATIONS 7 Long run economic and land use impacts of major urban infrastructure initiatives 7 Strategic implications 11 City shaping power 11 Investment appraisal tools 11 Investment decision making processes 12 Infrastructure funding 13

1 INTRODUCTION 15 1.1 Project background 15 1.2 Project objectives 16 1.3 Project approach and report structure 16

2 WIDER POLICY AND ECONOMIC ENVIRONMENT 18 2.1 Introduction 18 2.2 Melbourne’s planning history 18 Plan for General Development (1929) 18 Melbourne Metropolitan Planning Scheme (1954) 19 Planning Policies for Metropolitan Melbourne (1971) 21 Metropolitan Strategy Implementation (1981) 22 Victoria Capital City Policy (1994) and Living Suburbs (1995) 23 Melbourne 2030 (2002) and Melbourne @ 5 Million (2008) 24 2.3 Macroeconomic conditions 25 2.4 Considerations 28

3 PROJECT HISTORIES AND CONTEXTS 30 3.1 Introduction 30 3.2 The City Loop 31 Initial project drivers and anticipated impacts 31 Funding and management 31 Key success drivers and anecdotal impacts 33 3.3 CityLink 35 Initial project drivers and anticipated impacts 35 Funding and management 36 3.4 Box Hill Activities Area development 37 Initial project drivers and anticipated impacts 37 Key success drivers 37 3.5 Western Ring Road 38 Initial project drivers and anticipated impacts 38 Funding and management 39

Confidential Copy 1

Anecdotal impacts 40 3.6 Key conclusions 41

4 METHODOLOGY 42 4.1 Rationale 42 Measuring accessibility through Effective Job Density 43 4.2 Key Variables 47 4.3 Approaches 47 Measuring growth in employment and households 47 Changes to labour productivity 49 Human Capital Enhancements 50 Value Chain Accessibility Benefits 51 Dwelling and Land Value Impacts 52 Walk time impacts 52

5 IMPACT ASSESSMENT 53 5.1 CityLink 53 Impacts to Employment and Housing 53 Economic Expansion 61 5.2 Western Ring Road 71 Impacts to Employment and Housing 71 Economic Expansion 78 5.3 The City Loop 89 Impact to Employment 89 5.4 Box Hill Central Activities Area development 93 Impacts to Employment and Housing 93 Economic Expansion 95

6 KEY CONCLUSIONS 99 6.1 Summary of findings 99 6.2 Areas for Further Refinement 101 6.3 Conclusions 102

APPENDIX 1: DATABASE DEVELOPMENT 103 Travel Time Matrices 103 Population and Dwellings 104 Employment by industry 105 Labour force participation 107 Labour productivity 108 Effective Job Density 110 Property values 111 Comparison of Box Hill and Preston Activities Areas 114

APPENDIX 2: TECHNICAL NOTES 117 Labour productivity estimation 117 Accessibility model 122 Agglomeration Model 125 Human capital benefits estimation 134

REFERENCES 141

Confidential Copy 2

Figures

Figure 1. Population Growth Trajectories ...... 9 Figure 2. Benefit Stream Across Time $Billion (Excluding Box Hill) ...... 10 Figure 3. Taxation Revenue as a Proportion of GDP, Ordered by Tax Burden, 2003 ...... 14 Figure 4. Melbourne Central Business District (1954) ...... 19 Figure 5. Northland Shopping Centre (1971)...... 20 Figure 6. 1969 Transport Plan – Road and Rail Network Plan ...... 21 Figure 7. 1971 Corridor Growth Plan ...... 22 Figure 8. Melbourne Skyline (1981) ...... 23 Figure 9. Southbank Skyline (1996) ...... 24 Figure 10. Melbourne 2030 Activity Centres Hierarchy ...... 25 Figure 11. GDP Growth Rates Volume Measure ...... 26 Figure 12. Annual Average Unemployment Rate...... 27 Figure 13. GDP Per Capita Growth...... 27 Figure 14. Selected Indutries Share of GDP ...... 28 Figure 15. Congestion at Flinders Street Station ...... 31 Figure 16. Congestion at Flinders Street Station ...... 32 Figure 17. CBD Development in the 1950’s ...... 33 Figure 18. Employment Numbers by CBD Block (1954)...... 34 Figure 19. Early City Ring Road and CBD Access Routes (1954)...... 35 Figure 20. Citylink Western and Southern Components ...... 36 Figure 21. Planned Box Hill District Centre (1954)...... 37 Figure 22. Proposed Arterial Road Network (1954)...... 39 Figure 23. Western Ring Road Connectivity ...... 40 Figure 24. Demand for Different Land Use as a Function of Accessibility ...... 42 Figure 25. Relative EJD Regression Coefficients...... 45 Figure 26. Household Share Regression Coefficients ...... 46 Figure 27. Accessibility Modelling Approach ...... 48 Figure 28. Agglomeration Method – Labour Productivity ...... 50 Figure 29. Agglomeration Method – Human Capital ...... 51 Figure 30. Citylink Impact to Effective Job Density, 2011 ...... 53 Figure 31. Employment Impacts, Citylink ...... 54 Figure 32. CityLink Employment Impacts, 2011 ...... 55 Figure 33. CityLink Transport & Wholesale Trade Employment Impacts, 2011 ...... 56 Figure 34. CityLink Knowledge Intensive Employment Impacts, 2011 ...... 57 Figure 35. CityLink Population Serving Employment Impacts, 2011 ...... 58 Figure 36. Citylink Household Impacts, 2011 ...... 59 Figure 37. Citylink Agglomeration Benefits, 2011 ...... 63 Figure 38. Citylink Additional Employment Benefits, 2011 ...... 64 Figure 39. Citylink Human Capital Benefits, 2011...... 66 Figure 40. Citylink Impacts to House Prices, 2011 ...... 69 Figure 41. Citylink Impacts to Unit Prices, 2011 ...... 70 Figure 42. Western Ring Road Impacts to Effective Job Density, 2011 ...... 71 Figure 43. Employment Impacts, Western Ring Road ...... 72 Figure 44. Western Ring Road Employment Impacts, 2011...... 73 Figure 45. Western Ring Road Transport & Wholesale Trade Employment Impacts, 2011 ...... 74 Figure 46. Western Ring Road Knowledge Intensive Employment Impacts, 2011 ...... 75 Figure 47. Western Ring Road Population Serving Employment Impacts, 2011 ...... 76 Figure 48. Western Ring Road Household Impacts, 2011 ...... 77 Figure 49. Western Ring Road Agglomeration Benefits, 2011 ...... 81 Figure 50. Western Ring Road Additional Employment Benefits, 2011...... 82 Figure 51. Western Ring Road Human Capital Benefits, 2011 ...... 84 Figure 52. Western Ring Road Impacts to House Prices, 2011 ...... 87 Figure 53. Western Ring Road Impacts to Unit Prices, 2011 ...... 88 Figure 54. Walking Accessibility – City Loop vs no City Loop ...... 92

Confidential Copy 3

Figure 55. Employment in Box Hill and Preston, 1996 to 2011 ...... 93 Figure 56. Centres Share of SLA Employment ...... 94 Figure 57. Average Annual Growth Rate of Households, Preston & Box Hill ...... 95 Figure 58. Box Hill Impacts to Effective Job Density, 2011...... 96 Figure 59. Box Hill Additional Employment Benefits, 2011 ...... 98 Figure 60. Benefit Stream Across Time $Billion (Excluding Box Hill) ...... 100 Figure 61. Summary of Project Benefits in 2011, ($Million) ...... 101 Figure 62. Population in Melbourne and Victoria, 1981 to 2011 ...... 104 Figure 63. Historical Employment in the CBD and City of Melbourne ...... 106 Figure 64. Metropolitan Melbourne Employment by Industry, 1996 and 2011 ...... 107 Figure 65. Annual Participation Rate, Melbourne ...... 108 Figure 66. Melbourne Labour Productivity...... 109 Figure 67. Metropolitan Melbourne Industry Labour Productivity, 1996 and 2011 ...... 110 Figure 68. EJD in Melbourne, 2011 ...... 111 Figure 69. Sale Trends for Houses by Ring, 1992 to 2009 ...... 112 Figure 70. Median House Prices and Distance From CBD...... 113 Figure 71. Box Hill Hospital 1960s and PANCH Hospital 1950s ...... 115 Figure 72. Northland Shopping Centre, Myer Entrance 1967 ...... 116 Figure 73. SLA Professional, Scientific & Technical Services Labour Productivity ...... 122 Figure 74. Accessibility Modelling Approach ...... 123 Figure 75. Agglomeration Method – Labour Productivity ...... 126 Figure 76. Worked Example of Converting a Travel Time Matrix to SLA Matrix ...... 128 Figure 77. Worked Example of EJD Calculation ...... 129 Figure 78. EJD in Melbourne...... 129 Figure 79. Scatter Plot Labour Productivity and EJD, Professional Services ...... 131 Figure 80. Industry EJD Elasticities ...... 131 Figure 81. Melbourne Example Percentage Change in EJD ...... 132 Figure 82. Worked Example of Labour Productivity Impact on GVA ...... 134 Figure 83. Agglomeration Method – Human Capital ...... 135 Figure 84. Gross Annual Income Per Capita, Male Bachelor Degree ...... 136 Figure 85. Bachelor Degree Human Capital Stock, Males Aged 40 to 44 ...... 138 Figure 86. Worked Example of Human Capital Benefit...... 140

Confidential Copy 4

Tables

Table 1. Summary of benefits to Metropolitan Melbourne, 2011 ($million) ...... 8 Table 2. Scope of Benefits in (Transport) Infrastructure Cost Benefit Analyses (CBA’s)...... 12 Table 3. Relative EJD Index, Selected SLAs, 1996-2011 ...... 44 Table 4. Key Variables and Estimation Methods ...... 47 Table 5. CityLink Impacts to Employment and Households ...... 54 Table 6. CityLink Employment Impacts by SLA, 2006 and 2011...... 60 Table 7. CityLink Population Impacts by SLA, 2006 and 2011...... 60 Table 8. CityLink Impacts to Melbourne Gross Value Added ($000’s) ...... 61 Table 9. Sensitivity analysis of additional employment benefit, citylink ($000’s) ...... 62 Table 10. Citylink Impacts on Industry Gross Value Added ($000’s), Melbourne...... 62 Table 11. Citylink Gross Value Added Impacts by SLA, 2006 and 2011 ($000’s) ...... 65 Table 12. Citylink Human Capital Impacts to gva, 2006 and 2011 ...... 66 Table 13. CityLink Value Chain Accessibility Benefits ($million) ...... 67 Table 14. CityLink Value Chain Accessibility benefits to industry gross value added ($million) ...... 68 Table 15. CityLink land value impacts ...... 69 Table 16. Western Ring Road Impacts to Employment and Households ...... 72 Table 17. Western Ring Road Employment Impacts by SLA, 2006 and 2011 ...... 78 Table 18. Western Ring Road Population Impacts by SLA, 2006 and 2011 ...... 78 Table 19. Western Ring Road Impacts to Melbourne Gross Value Added ($000’s) ...... 79 Table 20. Sensitivity Analysis of Additional Employment Benefit, Western Ring Road ($000’s) ...... 79 Table 21. Western Ring Road Impacts on Industry Gross Value Added ($000’s), Melbourne ...... 80 Table 22. Western Ring Road Gross Value Added Impacts by SLA, 2006 and 2011 ...... 83 Table 23. Western Ring Road Human Capital Impacts to GVA, 2006 and 2011 ...... 83 Table 24. Western Ring Road Value Chain Accessibility Benefits ($Million)...... 85 Table 25. Western Ring Road Value Chain Accessibility Benefits, Industry GVA ($Million) ...... 85 Table 26. Western Ring Road Land Value Impacts ...... 86 Table 27. Impact of City Loop 2011, Employment Scenarios ...... 89 Table 28. Benefits of City Loop, 2011 ($Billion)...... 90 Table 29. City Loop Detailed Benefits for Density Scenario, 2011 ($Billion) ...... 90 Table 30. City Loop Land Value Impacts ...... 91 Table 31. Box Hill impacts on Industry Gross Value Added ($), Melbourne ...... 97 Table 32. Summary of Benefits to Metropolitan Melbourne, 2011 ($Million)...... 101 Table 33. Selected SLAs Average Annual Growth Rates of Population ...... 105 Table 34. Selected SLAs Average Annual Growth Rates of Total Employment ...... 107 Table 35. Labour Productivity by Region ...... 109 Table 36. Land Value Regression Results ...... 114 Table 37. Sector Classifications ...... 118 Table 38. Major Capital Cities Labour Productivity ($ of Gross Value Added Per Hour Worked) ...... 119 Table 39. Selected SLA Labour Productivity ...... 121 Table 40. Example Travel Time Matrix for Private Vehicle, Minutes ...... 126 Table 41. Estimate of Melbourne’s Human Capital ($ Billions) ...... 136

Confidential Copy 5

KEY TERMS

Term Definition

ABS Australian Bureau of Statistics

Agglomeration A term used to describe the economic benefits firms gain from locating near one another. Given the economies/benefits positive relationship between employment density and labour productivity, increasing employment density will generate higher output per worker. This translates to an overall increase in GDP. It is different, separate and additional to the M2MPJ benefit as it relates to workers who remain in the same job and not additional workers. CBD Refers to the Central Business District of Melbourne.

Effective Job Density EJD is a measure used to examine a location’s access to jobs and services across a wider metropolitan (EJD) area. The EJD of an area is the sum of employment in the area and the employment in all other areas divided by the travel time in reaching these external jobs, with travel time weighted for transport mode. Gross Value Added A measure in economics of the value of goods and services, excluding costs of production, produced in (GVA) a defined area, industry or sector of the economy. It is mainly comprised of income and business profits linked to productivity. This measure does not account for tax, but when added to taxes minus subsidies on profits it produces gross domestic product (GDP). Human Capital The labour, skills and knowledge embodied within a labour force which generate economic value. It can be measured as the lifetime earnings potential of an individual which can be affected by a variety of factors. One factor is the access to employment opportunities in a defined geographic region. LGA Local Government Area

Melbourne Refers to Metropolitan Melbourne, as currently defined by the Urban Growth Boundary or Melbourne Statistical Division. Move to More Relates to the increase in Gross Domestic Product (GDP) arising from additional employment occurring Productive Jobs in the city centre. Given the positive relationship between employment density and labour (M2MPJ) productivity, the increase in jobs in the CBD will generate higher output per worker. This translates to an overall increase in GDP. Reducing the capacity constraint on commuter trips into the city centre will allow more people to work in these highly productive jobs, thus increasing economic growth. SLA Statistical Local Area

Confidential Copy 6

SYNTHESIS AND STRATEGIC IMPLICATIONS

Long run economic and land use impacts of major urban infrastructure initiatives

This report provides an assessment of the long run economic and land use impacts of four infrastructure projects in Melbourne: the City Loop, CityLink, Box Hill Activities Area and the Western Ring Road. These projects have fulfilled their original, narrowly defined warrants relating typically to congestion busting and travel time savings. However, their impacts in shaping urban development and strengthening Victoria’s economy go far beyond those solely associated with transport improvements. The wider benefits include the:

- new jobs they have allowed Melbourne to generate - new households they have allowed Melbourne to accommodate - labour productivity improvements they have allowed - increase in human capital accumulation they have encouraged - transformation of value chains because of greater speed and certainty in freight movements.

There is a significant amount of literature and empirical evidence which outlines the theoretical underpinnings of these economic benefits. However, it should be noted that conceptualising and quantifying the economic benefits of transport projects is a new field. As more projects are appraised and the methodology is refined and debated within the transport appraisal and cost benefit communities there is likely to be ongoing refinements to the methodology.

The benefits generated by each of the four selected projects is summarised in Table 1. 2011 has been provided as an example of the long run impacts. The measure used to calculate these benefits throughout report is connectivity, defined by effective job density. This is a measure developed by SGS which quantifies access to employment.

The improved accessibility offered by CityLink has facilitated a further 70,300 jobs to locate in Melbourne along with an extra 58,200 households compared to a base case without this infrastructure. The project added $8.94 billion to the Melbourne economy during 2010-11 (3.8 per cent of Melbourne’s GDP) via employment growth and labour productivity improvements. In addition, the reduced freight costs due to the project resulted in GDP increasing by $81 million (0.03% of GDP). The increase in human capital allowed a further increase in GDP of $14 million (0.01 per cent of GDP).

The City Loop’s impacts in 2011, versus the base case, were of a similar order to CityLink’s. The Western Ring Road also had a substantial impact generating a gross value added (GVA) uplift of more than $2.5 billion and creating 24,900 net new jobs.

The Box Hill Activities Area has not contributed directly to increased employment in Melbourne, but the improved clustering of employment has increased GDP by $587 million (0.2% of Melbourne’s GDP).

Confidential Copy 7

TABLE 1. SUMMARY OF BENEFITS TO METROPOLITAN MELBOURNE, 2011 ($MILLION)

Benefit CityLink Western Ring Road Box Hill City Loop

Project Conceptualisation 1969 1954 1954 1929 Project Completed 2000 1999 1992 1985 Productivity Improvements $1,395 $228 $0 $1,205 Move to More Productive Jobs $7,547 $2,216 $587 $9,214 Total GVA Uplift $9,023 $2,593 $587 $10,419 New Jobs 70,300 24,900 0 74,000 New Households 58,200 17,400 n.a. n.a. Freight Improvements $81 $149 n.a. n.a. Freight Travel Time Savings -0.8% -1.6% n.a. n.a. Human Capital $14 $5 $0 n.a. Land Value Improvements $29,646 $10,174 $0 $21,139 Source: SGS Economics & Planning

These findings show that each project has provided a different mix of benefits to the economy of Melbourne. This mix appears to be related to the scale of the project and the economic geography which it has helped shape.

The City Loop has transformed the northern section of the CBD, while the north west of Melbourne has been opened up for both industrial and residential development by the Western Ring Road. CityLink has provided heightened connectivity between the CBD and north and south eastern parts of Melbourne. The data, statistical analysis and anecdotal evidence all support this.

The research findings suggest that the projects allowed the overall economy of Melbourne and Victoria to expand. In the early 1990s, before these projects commenced or began to generate significant dividends for the community, the ABS was anticipating a Victoria with 5.0 million people in 2011. In fact, over 5.6 million were resident in the State at 2011. While a range of factors contribute to population growth, the four infrastructure projects analysed helped to provide important conditions which allowed the economy to take a new growth trajectory. The infrastructure facilitated employment growth which required more labour (and hence more population) to fill these jobs. A larger population would then have had a stimulus effect on the economy, thus creating more growth.

Population projections provide an indication of the future growth trajectory of a region. In a developed economy such as Victoria, much of the population growth can be attributed to migration of working aged (25-54 years) people. Working-age migrations are drawn to a region by employment opportunities. Population projections developed by the Australian Bureau of Statistics in the early 1990s and published in 1996 had the population of Victoria increasing from 4.5 million to 4.9 million in 2011. Improved economic conditions above what had been expected in the early 1990s resulted in over 5.6 million people living in Victoria in 2011 (see Figure 1). There are a range of factors which could be associated with this higher than expected population growth. These could include the introduction of the baby bonus, increased popularity of Victoria as an export education destination and social changes reversing the downwards trend in fertility.

The improved economic performance of Melbourne would also have to be a clear factor. Infrastructure such as the City Loop, CityLink and the Western Ring Road opened up a range of economic opportunities. These projects are estimated to have enabled the Victorian economy to generate an additional 169,200 jobs. This extra growth could be seen as explaining around one third of the higher population growth than expected by the Australian Bureau of Statistics population projections.

Confidential Copy 8

FIGURE 1. POPULATION GROWTH TRAJECTORIES

Source: ABS Estimated Resident Population 2011 and ABS Population Projections, 1996

The benefits flowing from the three transport projects since their opening are shown in Figure 2. CityLink had a rapid ramp up in its benefit stream. The project improved the accessibility of large sections of Melbourne. A large number of small scale economic development events were able to occur fairly quickly across a large area within a diverse economy. The last three or four years have been impacted by more subdued background activity in the economy so it is difficult to ascertain if the growth in benefits from CityLink has reached a plateau or is merely experiencing a pause resulting from broader trends.

The period since the opening of the City Loop has seen a number of broad swings in background economic activity and this is reflected in the project’s flow of benefits. In Figure 2, 1991 to 1995 (years 11-15) correspond with the recession of the early 1990s and 2000 to 2003 (years 20-23) correspond with the economic slowdown following the introduction of the new tax system in 1999-2000. Those two plateaus also correspond with significant development occurring in Southbank (in the early 1990s) and Docklands (early 2000s). However, following these pauses the availability of the City Loop allowed ongoing private sector investment in the central city economy, generating a range of benefits. The most recent years have been impacted by slowing economic growth in the City of Melbourne.

Confidential Copy 9

FIGURE 2. BENEFIT STREAM ACROSS TIME $BILLION (EXCLUDING BOX HILL)

Source: SGS Economics & Planning

The Western Ring Road exhibits a similar pattern of impacts to CityLink but the ramp up period is not as pronounced. This could reflect less pent up demand or latent opportunity to generate economic activity compared to CityLink. CityLink was constructed through an established part of Melbourne, meaning it could address existing demand. The Western Ring Road on the other hand generated new demand through its greenfield location, facilitating growth of industrial and residential areas in Melbourne’s west.

The levelling off in benefits has been more significant for the Western Ring Road over the past few years. This may be associated with the capacity constraints in the original design of this infrastructure.

The profile of benefits shown in Figure 2 is also driven by the different roles of the Ring Road and CityLink within Melbourne. As the names suggests, CityLink helped integrate the north and south east of Melbourne with the central core of the metropolis, while the Western Ring Road has a more limited function, serving the industrial zones of the north west of Melbourne.

The economic development facilitated by these projects enabled employment, population and income growth to be significantly higher than the trends that had been anticipated prior to commissioning. This higher growth provided all levels of government with hundreds of millions of dollars more in taxation revenue than would have been the case without the infrastructure projects. It is of significant note that with the City Loop, government has enacted mechanisms designed to capture some of the uplift in land value which would flow from the project to assist with funding.

All the benefits which flow from transport projects should be considered when planning future infrastructure investments. However, traditional cost benefit analysis fails to capture the full gamut of benefits that city-shaping infrastructure can provide to an economy. Indeed, during their inception some of the projects analysed had a benefit cost ratio below one. The length of benefit streams, the appropriate discount rate and exclusion of wider economic benefits mean that traditional cost benefit analyses (CBAs) often do not support new infrastructure projects. This and related matters of project formulation and evaluation are considered further in the next sections.

Confidential Copy 10

Strategic implications

The research has a number of important implications for the way major infrastructure projects are planned and delivered in Victoria and for urban policy generally. These relate to the:  city shaping power of large scale infrastructure projects  adequacy of investment appraisal tools applied by government  opportunities to improve investment appraisal processes within government  funding of major infrastructure projects.

City shaping power The research has emphatically endorsed arguments developed by Eddington (2009), Graham (2006) and others that major infrastructure projects can, quite literally, re-sculpt the pattern of metropolitan development. Substantial shifts in the accessibility contours applying over urban space will change the location choices of firms and households alike, setting in place a new geography of land values. This, in turn, will signal where new and/or intensified urban development is warranted under commercial market rules. The outcome is a shift in urban form and, sometimes, structure.

These processes have been well rehearsed in the theoretical literature for decades. They have been intuitively understood in policy circles and within the wider community – people see, for example, the nexus between highway development and increasing land values and housing development in peri- urban regions. But, the mechanics and lags involved are under-researched. Moreover, the lack of quantitative evidence on this city shaping effect has slowed the full integration of transport and land-use planning in Victoria, notwithstanding that this State is probably one of Australia’s leading exponents of such integration.

The message from this report is clear; major transport investments are a powerful and, perhaps, the pre- eminent policy lever for determining metropolitan structure. Land use regulation via planning schemes and the like are more likely to play a supplementary role in managing urban development. This means that major transport projects need to be conceptualised within the context of a preferred urban structure, that is, ‘creating the sort of city we want’ as opposed to following the more conventional ‘predict and provide’ philosophy where transport investment simply responds to demonstrated demand. In some instances, it may make more economic sense to prioritise transport infrastructure that will reshape the city in permanently advantageous ways, over those projects that are solving evident congestion problems.

Investment appraisal tools A key investment appraisal tool applied within government is cost benefit analysis. This seeks to determine whether the stream of traded and external benefits generated by a project justify the capital, maintenance, operating and external costs involved, taking into account that the same resources could be deployed to other socially productive uses.

Cost benefit analysis, as applied to major transport investments, is evolving rapidly. The range of impacts taken into account in investment appraisal now extends well beyond direct user benefits and a limited range of environmental externalities (emissions, safety, neighbourhood disruption and amenity, etc). Agglomeration economies are increasingly taken into account in project appraisals, albeit in supplementary documentation to ensure full transparency of the impact of these factors on overall investment performance. Some appraisals are extending the scope of benefits further to take into account human capital enrichment (achieved by allowing households better connections to formal and informal learning opportunities) and the improvement to equity and social harmony which follows from allowing a greater palette of housing, employment, learning, health and recreational choices to households which currently have limited choices (see Table 2).

Confidential Copy 11

TABLE 2. SCOPE OF BENEFITS IN (TRANSPORT) INFRASTRUCTURE COST BENEFIT ANALYSES (CBA’S) Potential Benefits Generated by a New Transport Traditional CBA Traditional CBA + Traditional CBA + Link WEBs WEBs + Equity & Human Capital Effects Business transport costs are reduced, enabling    expanded production

Business to business synergies are improved (e.g.    economies of scale and scope)

Removal or mitigation of transport constraints on    the expansion of high value added industries in propitious locations

Labour participation and productivity are improved    as a result of reduced travel costs for workers and better labour matching

Human capital is enriched (expanded tactic learning    opportunities)

Households choice (consumption, learning,    employment) is expanded

Household travel costs are reduced   

Source: SGS Economics & Planning

The evidence and analysis in this report urges the early adoption by the Victorian Government of a broad spectrum of benefits in its investment appraisal tools. But even if a wide net is cast in terms of identified impacts, cost benefit analysis faces some structural weaknesses which limit its efficacy in the context of ‘city shaping projects’. It is conventional in cost benefit analysis to restrict measured impacts to the ‘first round effects’ of projects. These may be subject to lags, but have a direct ‘cause and effect’ link with the infrastructure item in question. Indirect and feedback effects are excluded, mainly for practical reasons; if second and subsequent round benefits are to be taken into account, so must costs, making the data gathering and analysis process very complex and open to challenge (because multiple judgements are required in identifying the effects).

The problem is that the profound city shaping power of major projects like CityLink or the City Loop is effected through a multitude of linked decisions where feedback loops are crucial. For these type of projects cost benefit analysis needs to evolve further, perhaps taking on more of the character of dynamic general equilibrium modelling, in terms of tracking feedback effects, and linking this to land use outcomes. This research begins to develop such methodologies.

Investment decision making processes It is clear from this research that a relatively small number of investment decisions on city shaping projects can set the scene for a large number of subsidiary or ‘follower’ investments in lower order roads, schools, hospitals and the like. Yet, there is little to distinguish city shaping and follower infrastructure in terms of project conceptualisation and formulation, as well as the depth of scrutiny and consideration, exercised during the State’s investment appraisal process. More expensive projects are certainly examined more closely, but this is not the point being made here; a billion dollar hospital decision is a substantively different one (with respect to the city shaper and follower dichotomy) than a billion dollar road which might significantly affect accessibility.

Confidential Copy 12

In principle, city shaping projects demand a more concerted ‘whole of government’ approach to planning and evaluation because they will have cross-portfolio implications for asset management and service delivery costs. There is currently good inter-departmental collaboration, between DPCD and DoT, in framing and testing such projects, but there may be scope for greater involvement by ‘economic’ agencies, including DBI.

Formalising the whole of government generation of city shaping projects and making this process more transparent to business and community stakeholders is likely to stand such projects in good stead when they are presented to Infrastructure Australia and the Commonwealth for funding. Bringing greater structure to the generation and evaluation of these projects is not to suggest that major investment is necessary in bureaucratic processes. The projects in question are few and far between and can be dealt with via time limited enquiry and co-ordination structures, rather than permanent institutions.

Infrastructure funding Another important research finding relates to the disconnect between those spheres of governance responsible for identifying city shaping projects and those controlling the funding for these projects. This is not a new discovery. Vertical fiscal imbalance has long been recognised as one of the potential weaknesses of the Australian federal system. As explained by Morris (2006)1, vertical fiscal imbalance has evolved in Australia, rather than reflecting the original principles in the federation agreement. The fiscal power of the Commonwealth has increased markedly at the expense of the States “as a result of decisions taken at times of historical crisis and a series of High Court rulings”. The federal government raises about 80 per cent of all government revenue, but requires only 61 per cent for its own purposes. Intergovernmental transfers currently account for 50 per cent of total State revenues, up from around 10 per cent in the 1950s.

Vertical fiscal imbalance raises two issues. Firstly, there is the question as to whether the overall size of the infrastructure funding pool available to Victoria is smaller than it otherwise would be as the Commonwealth is in a position to expand its ‘own purposes’ expenditure at the expense of the States. Secondly, and perhaps more important in the context of the current study, the State may have an incentive to identify projects which are more likely to attract Commonwealth support rather than projects which deliver the best economic, social and environmental outcome for Melbourne and Victoria. Traditionally this has tended to bias State ‘bids’ towards inter-regional highways and freight links against, say, metropolitan public transport projects. This may have been mitigated via the COAG agreements on improved metropolitan strategic planning and the creation of Infrastructure Australia with a mandate for more holistic and strategic evaluation of projects. Nevertheless the structural problems remain. To the extent that such investments facilitate more efficient and productive patterns of city development, it is the Commonwealth which is the major beneficiary in a fiscal sense; there is no system of tax revenue sharing in place under which the state governments proposing such city shaping investments reap some of the productivity dividend. In an ideal world, the State would have greater access to untied funds in the pursuit of more efficient urban management (see SGS, 2011)2.

The research also points to value capture as an efficient mechanism to help fund major infrastructure projects. Land value uplift in benefitted areas is a consistent and substantial attribute of all the projects covered, with the possible exception of the Box Hill ‘district centre’ designation, though even in this case, more detailed data analysis may have revealed a positive effect.

Value capture is routinely used around the world as an infrastructure funding strategy. It has been used in the past in Melbourne but on the whole Australia is unusual in its failure to tap this mechanism. While Australia’s overall tax rate, at 31.6 per cent of GDP, is in line with the weighted average for the

1 Morris (2006) Australia: Equity, Imbalance and Egalitarianism, in Blindenbacher, R & Karos, A. O. (eds) Dialogues in the practice of fiscal federalism: comparative perspectives (Forum of Federations). 2 SGS (2011) Funding Growth Areas in a National Urban Policy - Background Paper, prepared for National Growth Areas Alliance

Confidential Copy 13

OECD, the use of land taxes is less than the average and well below the levels applicable in the US, UK and Canada (see Figure 3 and Warburton and Hendy, 20063).

FIGURE 3. TAXATION REVENUE AS A PROPORTION OF GDP, ORDERED BY TAX BURDEN, 2003

Source: OECD Revenue Statistics, 2005, as quoted in Warburton and Hendy (2006)

3 Warburton, D. and Hendy, P (2006) International Comparison of Australia’s Taxes, Australian Treasury

Confidential Copy 14

1 INTRODUCTION

1.1 Project background

The recession of the early 1990s hit Melbourne’s economy more severely than the rest of Australia. Melbourne’s industrial base contracted sharply which had a range of flow on effects, including heavy migration out of Victoria. Today, Melbourne has emerged from the worst global recession since the Second World War in a strong position. So what has changed? The economic reform agenda of the past twenty years has helped transform Melbourne into a post-industrial knowledge intensive economy.

Macroeconomic policy settings and a microeconomic reform agenda carried out by both the Commonwealth and State Governments have greatly enhanced the Melbourne economy. The Hawke / Keating Government reforms of the 1980s (labour market reforms, floating of the dollar, etc), carried out with the fiscal and taxation initiatives of the early Howard years had a significant long term positive impact on Melbourne. In harness with National Competition Policy, the Kennett Government’s moves to commercialise, corporatise and even privatise certain infrastructure and State-owned businesses injected a fluidity and dynamism into the Melbourne economy.

Spatial policy reforms and key infrastructure projects also played a crucial role in Melbourne’s transformation into a leading knowledge intensive economy. Road projects such as the Western Ring Road, CityLink and EastLink helped to improve connectivity to and from the CBD, the Port of Melbourne and Melbourne Airport. The heavy rail capacity provided with the opening of the City Loop during the 1980s underpinned jobs growth and strong development within the Central Business District (CBD).

Government-inspired revitalisations of Southbank and Docklands provided the CBD with brownfield land to accommodate growth in knowledge intensive service industries. Other government investments at strategic locations along the primary public transport network (such as Box Hill and Dandenong) have also influenced the economic shape of Melbourne.

At the Victoria-wide level, the State Government’s investments in the Regional Fast Rail project took an important step towards greater integration between the labour market of Melbourne and those of Ballarat, Geelong and Bendigo. Improved accessibility between suburban industrial zones provided by road investment also allowed the development of an effective industrial supply chain.

All of these factors have helped to produce a diverse economy built on nimble, knowledge intensive firms and traditional industrial activities. There is no doubt that major infrastructure projects have increased the productivity and economic capacity of the Victorian economy. However, there has not been an assessment of the contribution of infrastructure projects to the economic capacity of Victoria. This research examines the various benefits which have flowed to Victoria from infrastructure investment and how they can be estimated and better understood. This evidence base will help inform how future infrastructure investment may impact on the long term economic prosperity of Victoria.

There is a significant amount of literature and empirical evidence which supports the evidence base underpinning the economic benefits of transport projects. However, it should be noted that conceptualising and quantifying these economic benefits of transport projects is still a new field. As more projects are appraised and the methodology is refined and debated within the transport appraisal and cost benefit communities there is likely to be on-going refinements to the methodology.

Confidential Copy 15

1.2 Project objectives

The objective of this research is to:

“document the long run impacts that key infrastructure projects have had on Melbourne’s land use and transport networks, the Victorian economy, metropolitan and regional labour markets and the industrial profile of Melbourne” (p. 4).

Source: Request for Quotation (RFQ) from the Department of Transport (DOT)

The RFQ initially nominated four major infrastructure projects:

 “Underground City Loop  CityLink  Box Hill Central Activity Area development and either  EastLink, the Western Ring Road, or Regional Fast Rail” (p. 4).

Thus a combination of rail, road and land use development projects is of interest to the DOT. Furthermore, DOT sought evidence of the land use and economic impacts of these projects on business costs; business clustering/agglomeration/competitiveness/innovation; access to markets; and labour force participation.

The City Loop, CityLink, Box Hill Activities Area and Western Ring Road were selected to be the focus of this report.

1.3 Project approach and report structure

Our approach is provided in the following flow diagram.

Inception meeting

Case study selection Selected commentator interviews

Macroeconomic conditions Data set development assessment

Initial impact assessment

Initial conclusions report

Detailed estimation process

Draft final report

Draft final report presentation

Final report

Confidential Copy 16

The report is structured as such:

Chapter 2: Wider Policy and Economic Environment Chapter 3: Project History and Context Chapter 4: Methodology Chapter 5: Impact Assessment Appendix 1: Database Development Appendix 2: Technical Notes

Within the report there are practical examples of business behaviour in response to the provision of infrastructure. These are presented within text boxes at various points throughout the report.

Confidential Copy 17

2 WIDER POLICY AND ECONOMIC ENVIRONMENT

2.1 Introduction

The way in which projects are conceived, planned and implemented reflects largely on the policy, political and economic environment of the time. In examining major infrastructure projects and their wider economic impacts, it is important to consider the policy and economic context to understand:

a) What were the unique drivers behind the success of the project and its impacts? b) If a similar project were undertaken today, would results be the same?

This chapter provides a historical overview of planning and the macroeconomic conditions which influenced how infrastructure projects have been planned and developed in Melbourne’s history.

2.2 Melbourne’s planning history

Founded in 1835 and declared a city by Queen Victoria in 1847, Melbourne as we know it today largely began with Robert Hoddle’s 1837 street layout, the ‘Hoddle Grid’. The grid, extending north-south from Flinders to Latrobe Streets, and east-west from Swanston to Spring Streets, provided Melbourne with a structured CBD. Melbourne’s first railway line, connecting the CBD with Port Melbourne was constructed in the 1850s, sparking a decade of rail building, where many of the metropolitan and regional railway lines in use today were constructed. These railway lines, privately funded and operated, were greatly extended during the land boom of the 1870s and 1880s. This period also saw the introduction of trams onto Melbourne’s streets, which connected new property estates with the existing railway network or CBD. While Melbourne started off as a tight, walkable city largely confined to the CBD and inner suburbs, railway and tramline investments allowed the city to greatly expand along transport corridors, facilitating extensive greenfield growth and the establishment of suburban Melbourne. As car use grew during the 20th century, locations not serviced by rail or tram began to be developed as the road and freeway network was expanded.

While the strategic infrastructure investments of the 19th century had directed the pattern of metropolitan growth, it was not until 1929 that Melbourne first published a strategic planning document, the 1929 Plan for General Development.

Plan for General Development (1929) By the early 1910s concerns about the dilapidated parts of the inner city and the rapid growth of the metropolis prompted major public inquiries by the Joint Select Committee on the Housing of the People in the Metropolis (1913-14) and a Royal Commission in 1915. Amendments to the Local Government Act in 1921 allowed Councils to implement zoning by-laws to establish residential districts, however holistic planning was delayed by World War (Freestone, 2008). It was not until 1929 that the Metropolitan Town Planning Commission produced the Plan for General Development. This planning scheme aimed to prevent ‘misuse’ of land and protect property values, highlighting rapid urban growth (3.5% per annum in the years leading up to the plan), rising traffic congestion, the distribution of recreational open space

Confidential Copy 18

and the haphazard development of land as key problems (Metropolitan Town Planning Commission, 1929). The plan forecast that the population of Melbourne would double over a twenty year period, increasing from around one million people in 1930 to around 2 million by 1950. At that point in time, Melbourne made up 56% of the population of Victoria. The plan had a large focus on the regulation of development, introducing zoning and calling for tighter design guidelines around road and other service provision, particularly in growth areas.

The onset of the depression followed by the Second World War saw little progress made in implementing the 1929 plan (Freestone, 2008; Department of Planning and Community Development, 2012). It was not until 1954 that the planning framework within which Melbourne largely operates in today was established with the creation of the Melbourne Metropolitan Planning Scheme.

Melbourne Metropolitan Planning Scheme (1954) The 1950s was another period of sustained population growth in Melbourne, with the population around 1.4 million and anticipated to rise to between 2.5 and 3 million people by the year 2000. Melbourne at this time comprised 60% of the Victorian population (Metropolitan Melbourne Board of Works, 1954).

FIGURE 4. MELBOURNE CENTRAL BUSINESS DISTRICT (1954)

Source: Metropolitan Melbourne Board of Works, 1954

Congestion in the CBD and inner areas and urban sprawl on the periphery of the city were again identified as key issues. As the report states Melbourne’s land was “not being put to the most appropriate use, public transport is loaded to strap-hanging capacity while the roads are so crowded that there is a constant danger to life and an economic loss through long delays and slow movement of traffic” (Metropolitan Melbourne Board of Works, 1954). Additionally, the Metropolitan Melbourne Board of Works (MMBW) was concerned with declining inner urban amenity, the need to expand industrial land supply in appropriate locations, a lack of available land for educational, medical and public space purposes, and a need to protect the city from the effects of aerial warfare. As such, the plan focused on:

 Limitation of the urban area  Zoning of specific areas for various community purposes  Decentralisation within the urban area of industry and commerce

Confidential Copy 19

 Provision for an adequate road communication system  Reservation of adequate areas for all community needs  Preservation of existing opportunities for civic improvement.

The plan was prepared on the basis of a future population of 2.5 million people contained in an urban area of about 265 square miles (686 km²). While the plan set to limit the outward expansion of the urban area and put in place zoning to ensure the timely provision of infrastructure to new development, it largely saw decentralisation as an inevitable force, one which could either be planned or unplanned. As such, the 1954 plan saw the emergence of five district business centres; Footscray, Preston, Box Hill, Moorabbin, and Dandenong envisaged to offer amenities comparable to those of the central business district. Such centres would be largely serviced by car and dependent on the road network, a marked step away from the previous model of urbanisation heavily dependent on access to the central city facilitated through train and tram connections.

The 1954 plan left a lasting legacy on Melbourne’s urban development pattern, and was influential in the transport studies and plans that emerged during the 1950s and 60s. However, while influential, the rise of free-standing shopping malls (Figure 5) and the extent of post war population growth were not fully anticipated in the 1954 Plan (Freestone, 2008). This led to a more corridor based growth pattern being adopted, reflected in the 1969 Melbourne Transport Plan, which proposed major freeway construction throughout the metropolitan area, as well as rail projects including the City Loop, Doncaster and Monash rail network extensions (Figure 6). A majority of the 1969 Transport Plans budget was devoted to road construction.

FIGURE 5. NORTHLAND SHOPPING CENTRE (1971)

Source: Metropolitan Melbourne Board of Works, 1971

Confidential Copy 20

FIGURE 6. 1969 TRANSPORT PLAN – ROAD AND RAIL NETWORK PLAN

Source: Metropolitan Melbourne Board of Works, 1971

Planning Policies for Metropolitan Melbourne (1971) The 1970s saw a slightly different direction taken in planning Melbourne’s urban development. In 1971 Melbourne had a population of around 2.4 million, accounting for 72% of the Victorian population (Metropolitan Melbourne Board of Works, 1971). The 1971 Planning Policies for Metropolitan Melbourne promoted urban corridors separated by protected green spaces. These green spaces, known as green wedges consisted of areas of significant landscape, historic and scientific interest, major agricultural resources, water catchments, and areas supporting significant bird, animal and plant life. This policy saw the urban area of Melbourne extended some 5,000 km² along the defined corridors (Figure 7).

Confidential Copy 21

FIGURE 7. 1971 CORRIDOR GROWTH PLAN

Source: Metropolitan Melbourne Board of Works, 1971

At the time it was acknowledged that a corridor orientated growth plan would require significant improvements in the accessibility of outer regions. Emphasis was placed on a balanced transport modal outcome, however, the plan highlighted that “the fact remains that if vehicle ownership continues to expand as rapidly as expected, it will still be necessary to carry out a substantial programme of freeway and arterial road construction” (Metropolitan Melbourne Board of Works, 1971). While large scale road construction was planned for in the 1969 transport and 1971 land use plans, only a fraction of the proposed urban arterials and freeways were constructed, due in part to the ‘freeway revolts’ of the early 1970’s.

Metropolitan Strategy Implementation (1981) Following the end of the ‘long boom’ in the mid to late 1970s, the planning direction established in the 1981 Metropolitan Strategy Implementation Plan took on a number of new objectives. In pursuit of a ‘more diverse, more interesting, more dynamic’ Melbourne, the 1981 plan called for a vigorous Central Melbourne, continued but slower growth in the outer urban areas and greater activity at identified district centres along existing transport routes (Metropolitan Melbourne Board of Works, 1981). Figure 8 highlights the relatively low level of development in the CBD at the time.

The Plan emphasised an incremental development approach, taking full advantage of the public and private investment that already existed in the metropolitan region, including transport, recreational and cultural facilities, social service networks and gravity feed services.

Confidential Copy 22

Revitalising the central city was of particular importance to the plan, with the MMBW committing up to a million dollars per year for the prompt delivery of required new services. The central city (including the CBD, St Kilda Road and Victoria Parade) was also rezoned to more easily facilitate commercial development and encourage intensified housing development (Metropolitan Melbourne Board of Works, 1981). The effort towards repopulating the central city was greatly boosted by the launch of the City of Melbourne’s ‘Postcode 3000’ policy. Beginning in 1992, this policy encouraged residential apartment uptake in the central city, reversing the long term flight of population from the inner urban areas.

FIGURE 8. MELBOURNE SKYLINE (1981)

Source: Metropolitan Melbourne Board of Works, 1981

In an attempt to spur district centre development, the MMBW proposed relaxing planning controls, allowing development bonuses for large scale sites, relaxing parking provision requirements to density and proximity to public transport, and providing assistance for land assembly. District centres were identified at Box Hill, Camberwell Junction, Cheltenham/Southland, Dandenong, Footscray, Frankston, Glen Waverley, Greensborough, Moonee Ponds, Oakleigh, Prahran (Chapel Street), Preston, Ringwood and Sunshine (Metropolitan Melbourne Board of Works, 1981).

The 1980s saw the introduction of the Planning and Environment Act (1987), the legal framework for the preparation and administration of local planning schemes. Under this system, major amendments to planning schemes that attract submissions are referred to independent panels, while disputes over the issuing or otherwise of planning permits is heard by the Victorian Civil and Administrative Tribunal (VCAT).

Victoria Capital City Policy (1994) and Living Suburbs (1995) In the early 1990s the MMBW, which had produced all of metropolitan Melbourne’s strategic plans since 1954, was disbanded. With the MMBW’s disbandment, planning responsibility fell within the control of the State Government. During the 1990s the Kennett Liberal government made several reforms to the planning system, primarily linked to economic prosperity and job creation. Generally, and perhaps

Confidential Copy 23

reflecting Victoria’s poor economic health at the time, the plans of the 1990s had a shorter timeframe than previous plans, with implementation periods of five years common. Following both the amalgamation of local councils and an increase in State Planning Minister ministerial powers, planning power was consolidated to the state level (Freestone, 2008). Planning in this period revolved around major projects and large site redevelopments, notably; CityLink, the Western Ring Road, and the opening up of formally industrial land for residential and commercial redevelopment in Docklands and particularly Southbank (Victorian Government and the Melbourne City Council, 1994).

FIGURE 9. SOUTHBANK SKYLINE (1996)

Source: Walking Melbourne

In the mid 1990s, Melbourne’s population was around 3.2 million, accounting for 71% of the Victorian population. It was forecast to grow by 440,000 people by 2011 (Victorian Government, 1995).

Melbourne 2030 (2002) and Melbourne @ 5 Million (2008) With a change in State government in 1999, the Bracks Labor government brought about a return to long term strategic land use planning, releasing Melbourne 2030 in 2002. The plan was to guide Melbourne’s development over a 30 year period, with the population projected to reach 5 million by 2030 (Department of Infrastructure, 2002). It strove to produce a more compact city through the prioritising of infill as opposed to greenfield development; aiming to reduce urban sprawl, build up neighbourhood centres and protect the green wedges of Melbourne. To encourage such a development pattern an urban growth boundary was established and numerous activity centres of various influence nominated. Out of centre growth was restricted with the intention of protecting the amenity of established urban areas (Department of Infrastructure, 2002).

Confidential Copy 24

FIGURE 10. MELBOURNE 2030 ACTIVITY CENTRES HIERARCHY

Source: Department of Infrastructure, 2002

However, in 2008 the State government revised Melbourne 2030 with the planning update Melbourne @ 5 Million. With revised population projections, Melbourne was anticipated to reach a population of five million much sooner than Melbourne 2030 had forecast, with the population projected to be more like seven million by 2030 (Department of Planning and Community Development, 2008). As such, Melbourne @ 5 Million extended the established urban growth boundary and reduced the number of activity centres to six. These centres, known as Central Activity Areas (CAAs), are intended to offer CBD- like services and functions. The CAAs are: Box Hill, Broadmeadows, Dandenong, Footscray, Frankston and Ringwood.

2.3 Macroeconomic conditions

Macroeconomic policy settings and the microeconomic reform agenda carried out by the Commonwealth and State Governments have greatly shaped Melbourne’s economy. The Commonwealth’s microeconomic reform agenda under the Hawke and Keating Governments along with the fiscal and taxation initiatives of the Howard years had positive economic impacts for Victoria, notwithstanding the 1991-2 recession in the state.

Exposing the Victorian economy to global markets (and a reduction in tariffs for example for textiles, clothing and footwear) with the floating of the exchange rate and introduction of foreign banks within Australia saw an ‘internationalisation’ occur. In 1992, governments agreed to establish the National Competition Policy, which led to a series of reforms, significantly recognised as making positive welfare contributions. A Productivity Commission Review of National Competition Policy Reforms in 2005

Confidential Copy 25

highlighted that the National Competition Policy delivered ‘substantial benefits’ to Australia. In conjunction with the reforms, the Kennett Government’s moves to commercialise, corporatise and even privatise certain infrastructures and State owned businesses injected a fluidity and dynamism into the Victorian economy.

Growth in GDP4 for Melbourne and Australia between 1975-76 and 2010-11 is shown in Figure 11. In the recession of the late 1980s and early 1990s the Melbourne economy was battered more extensively than the rest of Australia. The industrial heartland of the city contracted sharply as global demand for manufactured goods fell. This had a range of flow on economic effects on the broader economy. This recession was made more severe by high interest rates and by local events such as the collapse of the State Bank of Victoria. There was heavy migration out of the so-called rust-belt of Melbourne. Population growth fell to levels not seen since the depression of the 1930s. The unemployment rate (Figure 12) in Melbourne almost doubled between 1989-90 and 1990-91. By 1993-94 the unemployment rate in Melbourne was 1.1 percentage points higher than the Australian average.

FIGURE 11. GDP GROWTH RATES VOLUME MEASURE

Source: ABS cat no 5204 and SGS Economics & Planning

The economy slowly recovered during the early 1990s until the Asian Financial Crisis which occurred around 1997-98 had a small impact on economic growth. Melbourne experienced a larger boom in 1998- 99 and a larger bust in 2000-01 than the rest of the country. This period was influenced by the introduction of the new taxation system which caused changes in consumption patterns to avoid the Goods and Services Tax (GST). 2000-01 was also the time of a recession in the United States. In most years since 2000-01 the Melbourne GDP growth rate has been higher than the Australian GDP growth rate.

4 GDP (Gross Domestic Product) refers to Australia, GSP (Gross State Product) refers to a State, while GCP (Gross City Product) refers to a city. But for simplicity sake in this paper all different measures are referred to as GDP.

Confidential Copy 26

FIGURE 12. ANNUAL AVERAGE UNEMPLOYMENT RATE

Source: ABS cat no 6291.0

However, when examining the GDP Per Capita measure, Melbourne has not performed as well. In most years, GDP Per Capita Growth in Australia has been higher than in Melbourne.

FIGURE 13. GDP PER CAPITA GROWTH

Source: ABS cat no 5204.0 and SGS Economics & Planning

Confidential Copy 27

Melbourne’s economic fortunes over the past three decades have been underpinned by a transition from an industrial focused economy to one rich in knowledge intensive services. In the 1970s and early 1980s Manufacturing produced between 21% and 22.5% of all income generated in Melbourne. Since that time it has been in steady decline and in 2010-11 Manufacturing represented 8.4% of Melbourne’s income. Over the same period Financial & insurance services increased from around 4% in 1974-75 to 14.8% in 2010-11. Professional services overtook Manufacturing in 2010-11 as the second largest industry in Melbourne. It generated 8.5% of all income in Melbourne.

FIGURE 14. SELECTED INDUTRIES SHARE OF GDP

Source: ABS cat no 5220.0 and SGS Economics & Planning

2.4 Considerations

Over Melbourne’s existence planning policy and direction has shifted back and forth, however, some key principles have continued to emerge. Early Melbourne did not have a comprehensive metropolitan plan, but growth was heavily dependent on the provision of heavy train or tram infrastructure. In early Melbourne, transport infrastructure fundamentally influenced the urban growth pattern, with the planning of infrastructure close to a proxy for land use planning. With the 1929 plan Melbourne was given strategic direction disentangled from transport investments. Largely a land zoning plan, the 1954 plan impacted much more significantly on Melbourne’s trajectory, aiming to limit outward urban expansion while also building up sub-centres throughout the metropolitan region.

In the years that followed, while there have been alterations, the principles of limiting outward urban expansion, later articulated through an urban growth boundary, and the desire to decentralise some activities to regional centres, later known as activity centres/areas, has remained relatively constant. Perhaps the most notable difference between Melbourne’s various planning documents is the emphasis placed on the central city. Early Melbourne until the 1950s was heavily focused on central Melbourne. The 1950s saw a change in direction, with more intense decentralisation policy the preference. This trend extended until the 1980s at which point the health of the central city had deteriorated to a point where planning intervention was necessary. From the 1980s onwards the central city has returned to some of its pre-1950s eminence, capturing residential and commercial development that in the decades prior had been locating in the suburbs.

Confidential Copy 28

The historic planning and economic environment of Melbourne highlights some of the drivers of infrastructure projects. Importantly however, it also presents factors which need to be considered alongside an assessment of their impact. Indicators such as jobs growth can be affected by infrastructure projects and planning policy, along with Melbourne’s transformation into a leading knowledge-intensive economy. Spatial policy reforms and infrastructure projects have undoubtedly played a crucial role in this transformation.

The heavy rail capacity which had been provided with the opening of the City Loop during the 1980s also underpinned jobs growth and development within the Central Business District (CBD) by increasing the overall capacity of the network and increasing inner-city station catchments. Road projects such as the Western Ring Road, CityLink and EastLink helped to improve connectivity within the city. This improved business to business interaction, reduced freight costs, opened up suburban industrial zones and made viable a new range of housing. These projects enhanced and reshaped the city.

Confidential Copy 29

3 PROJECT HISTORIES AND CONTEXTS

3.1 Introduction

In reviewing past infrastructure projects, consideration must be given not only to the anticipated impacts on the time, but to their influence on the style and shape of Melbourne. Historic planning strategies have left a significant legacy on planning on the future, and in reviewing these strategies and through expert interviews, it is clear that there is a strong sense of repetition in the projects that arise. This is not necessarily through a lack of creativity, but a reflection of the strategic nature of major infrastructure investments, the lag time between idea and materialisation, and the ebbs and flows of political and cultural movements.

This retrospective appraisal of the City Loop, CityLink, Box Hill Activities Area and Western Ring Road uncovers the impacts (both anticipated and unforeseen), and sheds light on the history and dynamics at play in the planning and realisation of these projects, offering valuable insights and lessons for the future.

For each project this chapter of the report provides a summary of:  how the infrastructure projects came about/what their drivers were;  what the initial impacts were expected to be;  how the projects were funded and managed; and  how the projects affected city development and other conditions (property prices, population, etc).

The content provided in this chapter has been sourced through a review of literature, including journal articles and policy documents, and a series of interviews with academics, historians and experts5 who were involved with the projects, either through research or in their inception or implementation.

5 Interviewees include: Mike Berry (RMIT University), Murray Cullinan (Department of Transport), Jennifer Cunich (Property Council of Australia), Graeme Davison (transport historian), Geoff Lawler (City of Melbourne), Lyndsay Nielson (former secretary of DOI and DSE), and Rob Milner (former planner of Box Hill City Council)

Confidential Copy 30

3.2 The City Loop

Initial project drivers and anticipated impacts As early as 1929, the Metropolitan Town Planning Commission recommended the construction of an underground loop, linking the existing eastern and western railway systems, somewhere near North Melbourne and Richmond above-ground stations. At the time Flinders Street was one of the busiest railway stations in the world and suffering from high levels of congestion (Marshall, 2011). The depression and the war years saw interest in the loop wane, but it was renewed in the 1950s, when the Victorian Parliamentary Public Works Committee reported favourably on the provision of a loop. While the report and committee provoked considerable interest, action was not forthcoming until the 1969 Melbourne Metropolitan Transport Plan. In 1971, an Act of Parliament was passed and the Melbourne Underground Rail Loop Authority (MURLA) established to oversee the construction and operation of the loop. That same year, the first sod was turned by the transport Minister of the time, Vernon Wilcox.

FIGURE 15. CONGESTION AT FLINDERS STREET STATION

Source: Melbourne Metropolitan Planning Scheme 1954: Congestion at Flinders Street, p.109

An aspiration of constructing the loop was to relieve congestion at Flinders and Spencer Street stations, enhancing the capacity of the network and providing greater access to the northern areas of the central grid. There was however, disagreement on what the main purpose of the loop was. Some stakeholders, such as the City of Melbourne at the time, saw the loop as a solution to a metropolitan wide train circulation problem, allowing the efficient turning around of east and west bound trains. Interviews suggest that, the loop was also intended to maintain the accessibility that allowed for the agglomeration benefits of central Melbourne to materialise. Development had continued to occur beyond a convenient walking distance of the existing railway stations, and there was a concern that this would induce unsustainable levels of car use, particularly if there was continued construction of freeways and provision of off-street parking (Rogan, 2012). As such, the loop was expected to stimulate public transport use and reverse the long-term trend of declining patronage.

Funding and management The plan initially included four stations, but was cut to the present three by the elimination of a station under Latrobe Street. It proposed a design similar to that agreed upon in the 1969 Transport Plan, which was different from previous plans in that the loop ran under Latrobe Street instead of Lonsdale Street as it was initially proposed.

Melbourne’s City Loop was completed in stages, with Museum (now Melbourne Central) the first station opened in 1981, followed by Parliament (1983) and Flagstaff (1985). With the completion of the loop,

Confidential Copy 31

five stations around the CBD were accessible from metropolitan rail lines, and the immediate congestion issues of Flinders Street station congestion were relieved.

Several sources provided funding for the loop, including State Government which provided the bulk of the funds, complemented by a special rates levy applied by the City of Melbourne to its rates base. The levy was intended to last for 50 years, but was abolished several years early in 1995. The funding rationale conceived in the 1960s was that the whole of the metropolis would benefit to the extent of three fifths of the total project benefits, while property owners of the inner city would receive special benefits amounting to two fifths of the total project benefits.

FIGURE 16. CONGESTION AT FLINDERS STREET STATION

Source: Melbourne Metropolitan Planning Scheme 1954: Congestion at Flinders Street, p.109

As such, the recommendation was that the State Government should provide 60% of the funds, and the ratepayers of the inner city the remaining 40% (Rogan, 2012). Later the Metropolitan Melbourne Board of Works (MMBW) was included in the funding scheme, assigned 25% of the costs, with the City of Melbourne assigned 25%, and the State Government assigned the final 50%. However, in the 1970s the City of Melbourne argued that Council bearing 25% of the costs was beyond their capacity to pay and would induce cripplingly large loan servicing costs (Melbourne City Council, 1970).

Confidential Copy 32

FIGURE 17. CBD DEVELOPMENT IN THE 1950’S

Source: Melbourne Metropolitan Planning Scheme 1954: Development of the CBD in 1954, p.117

Anecdotally, cost benefit analysis was conducted prior to commitment to the loop, and resulted in a benefit to cost ratio of below one (indicating that for every dollar invested in the loop, less than a dollar of benefit would be retrieved). However, as one interviewee noted, compared with in the past, there is now more confidence that inner Melbourne works within a global economic imperative. To this end, building on the capacity of the city to absorb employment, residential and educational growth is critical to its global competitiveness.

Through its construction, the loop provided a quantum increase in the capacity of the rail network. This capacity allowed all metropolitan train services to run through the loop, although in the original planning of the loop not all lines (the Glen Waverley line for example) were intended to. However, the patronage increases from the mid 2000s exhausted the capacity of the loop and, coupled with the scheduling difficulties presented by express and country trains, resulted in the loop nearing operational capacity by the late 2000s.

Key success drivers and anecdotal impacts Along with the transport infrastructure development of the city loop were several key land use developments that contributed to the success of the loop and the growth of central Melbourne. Firstly, the relocation of rail stabling yards from the centre of the network, around both Flinders and Spencer Street stations, to the end of the respective metropolitan lines enabled land on the edge of the CBD precinct to be redeveloped. Train stabling at the centre of the network was no longer required, as the through running of metropolitan services through the loop resulted in trains no longer having to reverse direction from the platforms of Flinders Street station.

The second major land use initiative was the redevelopment of almost the entire city block above the Melbourne Central loop station. As one interviewee remembers, the CBD north of Bourke Street prior to the construction of the City Loop was an unattractive place, covered with two to three storey factories and warehouses, with the State Library and RMIT the only attractors. The Melbourne Central site itself was a collection of working factories before the State Government oversaw the redevelopment of the site. Known as the ‘Victoria’, the redevelopment of the site extended the retail core of the city to La Trobe Street and, at the time of completion, housed Australia’s largest office building (City of Melbourne,

Confidential Copy 33

1985). The construction of the loop and the redevelopment of the land above it transformed the northern end of the CBD, allowing suburban residents to directly access inner city shopping by rail. From conducted interviews, this transformation was perceived to be particularly significant in the growth of the City North education precinct and a major catalyst for residential growth within the CBD. However, one interviewee noted that it was unlikely that in the 1960s the level of demand for inner city living currently being displayed had been fully comprehended by the planners of the time.

FIGURE 18. EMPLOYMENT NUMBERS BY CBD BLOCK (1954)

Source: Melbourne Metropolitan Planning Scheme 1954: Employment numbers by block, 1954, p.123

Finally, at the time the City Loop was nearing completion, development of the St Kilda Road and Southbank precincts of Melbourne had begun, further centring the heart of city on the Yarra River. There was concern that with the completion of the loop, land use development would not necessarily follow, leading to an underutilisation of the significant infrastructure investment.

The City of Melbourne changed their planning scheme in the mid 1980s, incorporating a number of initiatives, including increased plot ratios, increased parking provisions and sewerage works, to stimulate development specifically around the new loop railway stations and the northern CBD. The City Loop is now an integral part of Melbourne’s transport system and CBD.

Confidential Copy 34

3.3 CityLink

Initial project drivers and anticipated impacts As early as 1929, a proposal for a southern bypass to the CBD was put forward, however, it was the transport planning of the 1950s and 60s where the concept of a southern and western city bypass (later to become CityLink) gained more ground. The 1969 Melbourne Transportation plan advocated for reservations and set aside sinking funds for the new inner-city freeway system.

FIGURE 19. EARLY CITY RING ROAD AND CBD ACCESS ROUTES (1954)

Source: Melbourne Metropolitan Planning Scheme 1954: City Ring Road and CBD access routes, p.198

With the completion of many of the planned freeways of the 1950s and 60s, by the late 1980s Melbourne was served by several freeways that terminated in its inner suburbs, generally five kilometres from the centre. Once traffic left these freeways, it was distributed onto local and arterial roads, resulting in gridlock and some roads of the city, notably King, Spencer and Swanston Streets performing major through road arterial roles. Some residential and urban streets were handling volumes up to 80% greater than their planned capacity (Russell, 2000). While a degree of this traffic was bound for the CBD, many of the trips were cross town, largely servicing the regions beyond the central city, but forced through the CBD as there was no bypass available. During the 1980s Melbourne City Council advocated for a western bypass for the City, but in the 1990s first under the Labor Government, and then under the Liberal Kennett Government, the concept of a city bypass was taken further, and the project was conceptualised as an east-west connector (Institute of Engineers Australia, 2002).

Interviews highlighted that while the inner core of Melbourne is dependent on mass public transport, the next ring of the city, and particularly the city’s air and sea ports, rely on major road access. This interaction of the inner core with its surrounds allows for a high value services economy to be supported by the adjacent freight and logistics industry. However, prior to the development of CityLink, access and utilisation of the Port of Melbourne and between Melbourne’s individual freeways was poor.

The proposal to build CityLink was first announced in May 1992 and received formal State Government approval in mid-1994. An Environmental Effects Statement was carried out in 1994 for CityLink, with the objectives of the link stated as reducing through traffic on inner city streets; improving environmental outcomes; optimising economic benefits while minimising financial costs; and improving access between

Confidential Copy 35

industry and the port, rail and airport facilities (VicRoads, 1994). In 1995 the Melbourne CityLink Authority was established to assist development, evaluate submissions, negotiate with contractors, recommend a contractor, facilitate relations between agencies, and ensure the project was delivered according to the Act (VicRoads, 2008).

During the early 1990s the Victorian economy was in a state of despair, and CityLink was seen as an opportunity to stimulate the economy both through the creation of construction jobs and the more efficient movement of people and freight. A major challenge of the project was in the development of automated tolling that allowed vehicles to maintain freeway speed. This had not been developed anywhere in the world prior to CityLink. The total value of the project was estimated at $1.5 billion in 1993 prices (Muhammad and Low, 2006). CityLink was built between 1996 and 2000, with stages opened throughout 1999 and 2000. The full link was operational from December 2000.

FIGURE 20. CITYLINK WESTERN AND SOUTHERN COMPONENTS

Source: Transurban,

Originally proposed as a city bypass route, the final design included the Exhibition Street extension that provided CBD access, via Batman Avenue to the Monash Freeway and CityLink. This link provided access over the rail yards east of Flinders Street station and allowed the Route 70 tram to be rerouted to dedicated tracks between the sporting precincts. Interviewees indicated that the City of Melbourne was a primary supporter of this extension, partly because it enabled traffic to enter and exit the city without the use of Swanston Street or local roads in Southbank. There was no mention of implicit land use changes as a result of this project, however, it is likely that it spurred on a greater density of development within its immediate vicinity and off-ramps.

Funding and management CityLink was funded through a Public-Private Partnership (PPP), the first of its kind in Australia, with the contract awarded in 1995 to Transurban. As part of the development of CityLink, existing roads were upgraded and expanded, and tolling points were added. Toll charges were applied to new sections of road, as well as to the Monash and Tullamarine Freeways, which had previously not been tolled. One interviewee observed that the pricing structure put in place by toll roads encouraged better accessibility for higher value activities.

Operation of the road was leased on a 34 year concession, with the lease expiring 14 June 2034, after which it will be transferred to the State (Infrastructure Partnerships Australia, 2006). The contract between the State Government and Transurban offers protections for both parties. Protections include

Confidential Copy 36

the ability for Transurban to make a claim against the State Government should the government do anything that could reduce vehicle numbers on CityLink. Although unsuccessful, Transurban commenced legal proceedings in 2001 against the State of Victoria over the construction of Wurundjeri Way. It was claimed that this free road decreased the revenue of CityLink. A similar case could be argued if the State Government was to build or increase the capacity of road or rail routes that parallel CityLink.

As noted at interview, the PPP agreement developed for CityLink had not been conducted in the past and the business case was largely unknown. The PPP model developed during the negotiations of CityLink, known as BOOT (Build, Own, Operate and Transfer) was subsequently adopted throughout Australia.

3.4 Box Hill Activities Area development

Initial project drivers and anticipated impacts Once a stand-alone town, Box Hill was absorbed into Melbourne as part of the eastward expansion of the city in the 1950s. In 1954, the Metropolitan Melbourne Board of Works (MMBW) designated Box Hill as one of five district centres for metropolitan Melbourne (the others being Footscray, Preston, Moorabbin and Dandenong). The centre, even prior to the transport improvements of the 1970s, was based around the train and bus interchange station, was an established shopping and office centre, and included a regional hospital and regional educational facility – Box Hill TAFE.

FIGURE 21. PLANNED BOX HILL DISTRICT CENTRE (1954)

Source: Melbourne Metropolitan Planning Scheme 1954: Box Hill District Centre, p.58

In the mid 1970s the Box Hill Modal Interchange project was launched by Minister of Transport. This project involved the elimination of the Station Street level crossing through lowering of the railway line, construction of a new station, integration of a suburban bus and taxi station, provision of extensive commuter car parking, and commercial redevelopment of the air rights above the new station (Perrott et al, 1975).

Key success drivers Prior to the Box Hill redevelopment, interviewees recalled a very congested centre with an emerging economy that was rare in its ability to attract small scale suburban office development. The congestion around the centre was intensified by the level crossings around the station, bringing about a need for

Confidential Copy 37

grade separation. With the sinking of the railway, an opportunity to sell the air rights over the station emerged, with the land owned by the Victorian Railways. The air rights were sold to developers Lustig and Moar, with the sale partly funding the grade separation works. One interviewee described Moar in particular as a powerhouse developer. Lustig and Moar constructed the shopping centre that sits above the station, while Council closed off several local streets to traffic, greatly activating the precinct.

However, interviewees note that there were numerous other factors that contributed to Box Hill’s success. These include the availability of development-ready land, a strong migrant shopping history, an innovative and pro-development Council, and an abundance of money available for investment following deregulation of the banking system. These factors all occurring at the same time gave Box Hill the opportunity to develop to the degree it did. The success of Box Hill as a polycentric centre was strengthened when the Australian Tax Offices decided to locate there in the late 1980s, preferring Box Hill over Moonee Ponds and Dandenong. It was noted that while the upgrading and sinking of the station was important, the improved frequency and scheduling of train services that accompanied the hard infrastructure investment was potentially of greater significance.

With the combination of success factors outlined above, interviewees highlighted that most of the development of Box Hill occurred over a five year period from the mid 1980s and was largely complete by 1992. While milder levels of growth have occurred since 1992, much of this growth is attributable to the nearby regional education and health facilities, and later, in the early 2000s, the extension of the Route 109 tram. This extension had the intention of further strengthening the centre and was modelled on the past success of the rail integration with the centre. There was anecdotal evidence that the tram route extension had spurred higher density growth along the tramway.

Finally, one interviewee noted that Box Hill’s success was connected to its ability to attract institutional investors. In attracting institutional investment, the level of government commitment in fixed infrastructure was critical. Institutional investors, unlike private developers, needed to feel confident in recovering the value of their investments in the future, which could be some years away. Confidence in the future price of assets was enhanced through the provision of fixed, hard infrastructure.

3.5 Western Ring Road

Initial project drivers and anticipated impacts Planning for the Western Ring Road, as part of a much larger Outer Ring Road, dates back to the 1954 Melbourne Metropolitan Planning Scheme and 1969 Melbourne Transportation Plan. The scheme proposed a city ring road as well as a series of concentric ring roads at various distances from the CBD. Initially the reservation was protected by interim development orders, but later through road reservations. It remained in many transport planning documents for several decades, but re-emerged on the agenda in the 1980s, largely as an alternative to major freight traffic congestion on Pascoe Vale Road, between the Hume Highway and Tullamarine Freeway (Rumpf and Kiss, 1994).

Confidential Copy 38

FIGURE 22. PROPOSED ARTERIAL ROAD NETWORK (1954)

Source: Melbourne Metropolitan Planning Scheme 1954: Proposed Arterial Road Network, p.97

Funding and management Construction of the Western Ring Road commenced in 1989, with initial works centred around Broadmeadows. Over the next decade various sections of the route were completed and opened, with the entire route completed by 1999. Given the relative youth of the ring road, several interviewees noted that the full economic, transport and land use impacts of the Western Ring Road were still evolving and maturing.

The Western Ring Road extends 28 kilometres from the junction of the Princes and West Gate Freeways in Laverton to Sydney Road/Hume Highway in Fawkner. Through this section the Ring Road connects to all of Melbourne’s western and northern highways, namely: the West Gate, Princes, Western, Calder, Tullamarine and Hume freeways (See Figure 23).

Confidential Copy 39

FIGURE 23. WESTERN RING ROAD CONNECTIVITY

Source: SGS Economics & Planning

Eastwards of the Hume Highway, the Western Ring Road joins the much shorter Metropolitan Ring Road (10 kilometres). Together these form the national M80 highway, with the easterly section terminating at Greensborough. Initially the Metropolitan Ring Road was intended to orbit the whole metropolitan area, cutting through Eltham, Templestowe and Warrandyte, and link up with the now constructed Eastern Ring Road (‘Eastlink’). The ‘missing link’ through Eltham, Templestowe and Warrandyte has never been constructed, and at present there are no tabled plans to complete the link. Historically there has been sustained community opposition to the construction of the ‘missing link’.

Anecdotal impacts The Western Ring Road connects the individual freeways that service Melbourne’s sea and air ports, and as such, the movement of freight is a primary function of the road. The road relieves freight traffic from Sydney Road, Pascoe Vale Road and Geelong Road. The completion of the Ring Road was anticipated to deliver major economic benefits to Victoria by linking up the national freight corridors with the Port of Melbourne and Melbourne Airport (VicRoads, 1994a).

The heavy freight use of the connection has spurred industrial growth along the length of the route, resulting in a redistribution of Melbourne’s industry, which was noted during interviews. In the late 1980s the suitability of the wharves (that now house the ‘Docklands’ development) for port functions were being investigated. The decision was made to not rebuild the wharves, and hence the slow birth of ‘Docklands’ began. The construction of the Western Ring Road in the 1990s allowed the existing industries of the Docklands to relocate to cheap industrial land with good access to the port. This relocation freed up the Docklands area for residential and commercial redevelopment. During this same

Confidential Copy 40

period, retreating industries from the gentrifying areas of Southbank, Richmond and Collingwood also often relocated to new premises along the Western Ring Road.

Funding for the Western Ring Road was provided by both State and Federal Governments. In 1974, the Federal Government established a National Roads Program. Initially Federal funds were only provided to interstate Freeways, but progressively the program was broadened to include Metropolitan roads of significance, largely to improve freight movements. The National Roads Program provided the bulk of the funds for the Western Ring Road, supplemented by State Government funding. Recently there have been substantial improvements made to the road in the order of $2 billion, including additional lanes, dedicated exits, and traffic management systems.

3.6 Key conclusions

The way in which each project was conceived, planned and implemented differed considerably. Most projects had years, if not decades, of planning behind them. Interestingly, the initial drivers of projects tended to focus on operational improvements, particularly from a transport engineering and planning perspective. However, over time it is clear that the improvements generated by these projects extend beyond time travel savings and improved accessibility. Increased development rates, a higher level of jobs and the flow-on streetscape, urban design and amenity improvements associated with heightened accessibility have followed.

It is difficult to distinguish between the benefits generated by these infrastructure projects specifically and those which have arisen over time. One way to conceptualise the benefits is to imagine the city without a City Loop, the City Link and the Western Ring Road, and Box Hill without the infrastructure investment it has received. One can only imagine that the CBD would be less developed, Melbourne’s economy would be less efficient due to travel costs, and that Box Hill would emulate some centres which did not receive as much investment.

A key feature across all these projects is recognition by those planning and implementing them that they would be of long-term, strategic importance to the city.

Confidential Copy 41

4 METHODOLOGY

The methodology used to estimate economic and land use impacts from the City Loop, CityLink, Western Ring Road and investments into Box Hill Activities Area is outlined in this section. The varying ages of the projects necessitates a slightly different approach to estimating benefits amongst the projects, mostly due to data availability. This chapter presents the rationale to the methodology employed, key data and variables, and specific approaches used to estimate impacts of individual projects. Appendix 2: Technical Notes at the end of this report provides substantial detail on how particular variables were estimated.

4.1 Rationale

Well established economic theory indicates that over time, firms will tend to locate closer to areas that improve their land use efficiency. Locating in areas with superior accessibility reduces transaction costs through ease of contact with suppliers and customers. Crucially, this also increases access to a skilled labour force. The theoretical underpinnings of the relationship between land price and accessibility can be seen graphically in Figure 24.

FIGURE 24. DEMAND FOR DIFFERENT LAND USE AS A FUNCTION OF ACCESSIBILITY

Source: SGS Economics & Planning

Whilst all firms prefer locations of high levels of accessibility their ability to pay for locational advantages will differ, as do their aggregate land use demands. Land use demands between industries and the collar of the workers (blue and white) generally differ based on the functioning of their industries.

Land use demands by service industries are small relative to other industries, such as manufacturing and wholesale trade. Those industries require large amounts of land, relative to service firms. Furthermore, land rents can be shared across multiple firms as office towers use the relatively cheaper option of expanding vertically, rather than having large parcels of land. This contributes to the ability of service firms to locate within the confines of a heavily dense area of employment and population such as the CBD, whereas Manufacturing and Wholesale trade tend to locate further away from highly dense areas for their larger land requirements.

Confidential Copy 42

Ways to improve accessibility of firms differ across industries based on their customer and supplier base. Generally, Manufacturing requires quality road infrastructure and as a consequence tend to locate closer to areas where they have access to major road networks. Both their suppliers and customers also tend to have a similar accessibility requirement. Therefore efficiencies can be gained for those industries when they locate closer to points of road infrastructure. However, this is not necessarily the case for service industries. Whilst all firms require timely access to their suppliers, employees and customers, their strategic ability to access those people differ based on the function of their business models.

Opportunities for access to employment apply in a similar way to households as they do for industries. People, over time, will adjust their residential location due to a wide range of factors, some of which are access to employment, education, essential services and recreation. Literature indicates that these choices tend to be constrained due to factors such as family and historical ties to a region or corridor.

For these reasons many people and families may tend to locate ‘within corridors’, rather than moving ‘across town’. However, when moves are made within this context the relative accessibility of the two areas is a key consideration.

These theoretical predications are illustrated by recent experience in Australian cities, particularly with respect to the Western Ring Road, CityLink, EastLink and Westlink (M7) in Sydney.

Measuring accessibility through Effective Job Density SGS has developed a measure of accessibility within a specified geographical region and the ability to access overall economic activity across the wider MSD, known as Effective Job Density (EJD) (see Appendix 2 for technical notes).

The change in EJD associated with an infrastructure or land use initiative is estimated via the same method used to calculate the base EJD. The difference between the base case EJD and a project case EJD will depend on the project inputs and data availability. Both projected travel times and the share of public transport use could be affected by a project. Additionally, the level of future employment in a particular location could change. A project case EJD can then be estimated using these altered inputs and compared to the base case EJD. A project case EJD can be estimated for several time periods, such as 2016, 2021 and 2031. These can then be compared to a base EJD estimated for the corresponding year.

Each of the changes to EJD inputs will have a differing impact on EJD. Increasing employment in a particular location in general has the largest impact on EJD. This is most likely to occur under a land use project which increases employment at a targeted location. A reduction in travel times (both public and private transport) has a significant impact on increasing EJD, but to a lesser extent than employment changes. Increasing the share of public transport use can have adverse impacts on EJD depending on the project. If a greater share of commuters are travelling on public transport which still has a longer average travel time than private vehicle, EJD will fall.

Quantifying changes to relative accessibility is undertaken through translating absolute EJD values across the four years of analysis (1996, 2001, 2006 and 2011) into a 0 to 1 index, i.e. a relative EJD index. The index is based on the SLA that has the highest EJD. This was found to be Melbourne – Inner consistently across all years of analysis. This provides a ceiling for the index, a score of 1. To provide a floor to the index the converse is performed. Consistently Yarra Ranges – Central was found to have the lowest EJD within metropolitan Melbourne, and is thus given a score of 0. Table 3 demonstrates the relative EJD index across a selected number of SLAs from 1996 to 2011. As shown, changes in the location of employment have affected the index over time, with declining EJD in middle and outer SLAs. At the same time, new transport infrastructure has shifted the EJD in some locations. The result of this is that SLAs have in general, declined from 2006-11 in terms of relative EJD compared to the Melbourne – Inner SLA which has seen the largest boost in employment over time.

Confidential Copy 43

TABLE 3. RELATIVE EJD INDEX, SELECTED SLAS, 1996-2011 SLA 1996 2001 2006 2011

Melbourne - Inner 1.00 1.00 1.00 1.00 Melbourne - S'bank-D'lands 0.77 0.82 0.82 0.77 Melbourne - Remainder 0.63 0.64 0.65 0.61 Knox - North-East 0.23 0.23 0.24 0.22 Knox - North-West 0.28 0.28 0.30 0.27 Knox - South 0.28 0.27 0.29 0.27 Monash - South-West 0.44 0.44 0.45 0.42 Monash - Waverley East 0.36 0.35 0.37 0.34 Monash - Waverley West 0.40 0.40 0.42 0.39 Gr. Dandenong - Dandenong 0.33 0.33 0.35 0.32 Gr. Dandenong - Bal 0.32 0.31 0.32 0.30 Casey - Berwick 0.17 0.17 0.22 0.18 Casey - Cranbourne 0.19 0.19 0.21 0.20 Casey - Hallam 0.26 0.26 0.28 0.26 Casey - South 0.09 0.09 0.10 0.09 Source: SGS Economics & Planning

The regression equation used to estimate the relationship between industry movements, accessibility and households is shown below.

௜ ൅ߚ଴ ൅ߝ݁ݎ݄ܽܵܪܪ௜ ൅ߚଶܦܬܧ௜ ൌߚଵܴ݈݁ܽݐ݅ݒ݁݁ݎݕ݄ܵܽݎݐݏݑ݀݊ܫ Where:

௜= Relative Effective Job Density for SLA iܦܬܧݐ݅ݒ݈ܴ݁ܽ݁ ௜ = The share of an industry’s metropolitan employment within SLA i݁ݎݕ݄ܵܽݎݐݏݑ݀݊ܫ ܪܪ݄ܵܽݎ݁௜= The share of metropolitan Households within SLA i ߚ଴= Constant term ߝ= Error term

Accessibility also induces locational change for households, given they can access a greater amount of services and employment opportunities.

The regression equation that estimates accessibility induced household changes is shown below.

௜ ൅ߚ଴ ൅ߝܮܥܷܦܵܯ݋݂݁ݎ௝ǡௌ௅஺ ൅ߚଶ݄ܵܽܦܬܧ௜ ൌߚଵܴ݈݁ܽݐ݅ݒ݁݁ݎ݄ܽݏܪܪ Where:

௜= Relative Effective Job Density for SLA iܦܬܧݐ݅ݒ݈ܴ݁ܽ݁ ܪܪ݄ܵܽݎ݁௜= The share of Households within SLA i ݄ܵܽݎ݁݋݂ܯܵܦܷܥܮ௜= The share of overall urbanised areas within SLA i ߚ଴= Constant term ߝ= Error term

When applying this statistical relationship to a project the key coefficient relates to relative EJD. This coefficient represents the increase in a location’s share of metropolitan Melbourne’s total households or employment if a location was to shift from being the least accessible location to being the most accessible location.

Figure 25 presents the relative EJD coefficient for each employment industry sector and households. This can be interpreted as follows: If the accessibility of a location increased from 0 to 1 it would increase its share of metropolitan Melbourne’s total housing stock by 0.02 (or 2%), for example.

Given much of the total housing and employment stock is already established, we see that the coefficient is relatively small (i.e. 2%). However, as a share of growth within a defined period this could represent a substantial proportion of total growth.

Confidential Copy 44

Over the past 15 years much of Melbourne’s housing growth has been focused on two broad regions, fringe growth areas (about half) or inner city Melbourne (about one fifth). Greenfield growth is predominately a result of large amounts of land supply from significant areas of land being rezoned as urban residential. While strong growth within the inner city can be attributed to a preference to be close to jobs, services and the city’s core (i.e. accessibility). As a result households have been estimated to have a strong propensity to relocate due to changes in accessibility.

In relation to a firm’s locational preference, it can be seen that as previously suggested the service sector and higher value added industries exhibit a higher preference for more accessible locations. Information media & telecommunications, Professional, scientific & technical services and Finance & insurance are some of the most susceptible industries to changes in accessibility.

Agriculture experiences the only negative coefficient with relative EJD. Agriculture uses typically require large and relatively cheap land parcels away from major centres. Furthermore, particularly within the fringe growth areas, farm land within the MSD has been rezoned and changed uses to housing or other employment.

For employment there is a secondary effect resulting from the estimated shifts in households. That is, many industries depend on a local population as either customers or skilled workers. Therefore, an increase in population in an area often results in a further increase in population servicing employment. Figure 26 displays the coefficients related to a change in household share for each of the employment industry sectors.

As expected industries that thrive on proximity to population, such as retail trade, health care, education & training and accommodation & food services all have a strong positive relationship with an induced change in households.

FIGURE 25. RELATIVE EJD REGRESSION COEFFICIENTS

Source: SGS Economics & Planning

Confidential Copy 45

FIGURE 26. HOUSEHOLD SHARE REGRESSION COEFFICIENTS

Source: SGS Economics & Planning

The effect of changes to accessibility is the cornerstone in the analysis of impacts for the four projects. The varying ages of the projects and subsequent issues in data availability and the stream of benefits over time mean that different approaches are required to understand how they have shaped Melbourne’s population, household and employment formation and distribution.

Confidential Copy 46

4.2 Key Variables

A series of variables which potentially demonstrate economic and land use impacts of the transport and urban infrastructure investments was firstly collated. Data on a series of variables across the metropolitan area from 1986 to 2011 at the smallest feasible geographic level (Statistical Local Area) was collected. These are summarised in the following table and are accompanied with the source and type of data which was available for use. Appendix 1 provides greater detail on estimation methods and data sources.

TABLE 4. KEY VARIABLES AND ESTIMATION METHODS Variable Broad Estimation Method Population and dwellings Estimated at an SLA level from 1981 to 2011. From 1991 to 2011, Estimated Resident Population (ERP) produced by ABS is used. From 1981 to 1991, the Suburbs in Time database, produced by the Department of Planning and Community Development (DPCD) is used at a suburb level. Total population in Melbourne and Victoria is estimated back to 1981 using Australian Historical Population Statistics (cat. no. 3105.0.65.001). Employment by industry Estimated from 1996 to 2011 using ABS Census data which has been interpolated for years between, and benchmarked against the Labour Force Survey. Melbourne (C) further adjusted using Census of Land Use and Employment (CLUE) data. From 1986 to 1996, the City of Melbourne Strategy Plan 1985 and the Suburbs in Time database was used. Labour Force Participation Informed via ABS Labour Force Survey Labour Productivity Estimated using gross value added (GVA) for an industry, divided by total number of hours worked in that industry.

Travel Time Travel Time Matrix based on the Melbourne Integrated Transport Model (MITM). Property Values Valuer General Data from 1992 onwards available. Effective Job Density Estimated using employment by industry compared to relative time taken to gain access to that employment, based on the mode split of those workers.

Land Value Impact Regression analysis to estimate relationship between house prices (as proxy for land values) and effective job density.

Human Capital Benefit Estimated via comparing EJD to human capital stock by age and qualification at a small area level. A human capital elasticity is calculated and compared to EJD by age and qualification, and then examined against changes to EJD associated with a project.

Of the variables listed, Effective Job Density, Land Value Impact and Human Capital Benefits have been most extensively used to delineate impacts of projects to Melbourne. The following section outlines the way in which variables have been used among City Loop, CityLink, the Western Ring Road and investments in Box Hill Activities Area.

4.3 Approaches

Measuring growth in employment and households CityLink and Western Ring Road To estimate the induced change in households and employment from a road project, detailed analysis of changes in accessibility was undertaken by SGS. With the implementation of this project in the region two key components related to an areas access to jobs and services will be impacted and subsequently estimated. These are:  Changes in travel times; and/or  Changes in mode share uses (Public Transport versus Private Vehicles)

Confidential Copy 47

It should be noted that for this analysis the change in the distribution of people and firms is purely based on the induced effects of improved accessibility. While not captured in this analysis, in addition to these impacts, there can often be one or more targeted intervention developments near key locations (i.e. new stations) that may further capitalise on the improved accessibility of the corridor.

This model quantifies the observed relationship (at 1996, 2001 and 2006) that population and employment (by industry group) levels have with the relative accessibility (weighted travel time) of relevant SLAs and the wider metropolitan area. In short, the model predicts where population and employment will gravitate given augmented accessibility. This method is outlined in Figure 27.

FIGURE 27. ACCESSIBILITY MODELLING APPROACH

Source: SGS Economics & Planning

For the CityLink and the Western Ring Road projects, impacts to employment and housing have been estimated using an accessibility model. The EJD index is used to estimate changes to employment and households arising from changes to relative accessibility. These changes represent an additional benefit to Melbourne rather than a redistribution of current employment within Melbourne. That is, the economy of Melbourne has grown larger / at a faster rate than would have been the case.

The difference in relative EJD for each SLA between the project and base case was applied to the results from the regression analysis. The induced changes to the number and distribution of households and employment by industry were then estimated. The population increase stemming from the projects was estimated via average household size. City Loop For the City Loop, a different approach was used. The road projects have improved the accessibility of large sections of Melbourne, while the City Loop had a direct impact on a small part of the city. The concentrated nature of the impact has been that the development it has facilitated has been of a far greater scale. At the same time however, the City Loop has boosted capacity in the train network, which has impacts throughout metropolitan Melbourne. It is difficult however, to delineate benefits stemming from increased network capacity caused by the City Loop compared to other capacity-enhancing investments, such as line duplication for example. As such, the focus has remained largely on the CBD.

A further complication in assessing benefits generated by the City Loop is the age of investment. The City Loop was opened over the 1980s, meaning that Travel Time Matrices and patronage data is not as readily available as compared to road projects which can be understood via transport modelling.

Because of data availability and difficulty in ascertaining the level of households which shifted due to the City Loop, employment is instead focused on. To estimate employment impacts, three techniques are employed:

1. Density approach. Differences in employment density within the CBD (Hoddle Grid) are used to highlight the impact the City Loop has had on development. Sites which have good access to

Confidential Copy 48

the three City loop stations (Melbourne Central, Flagstaff and Parliament) are compared to sites that have less access. 2. Share analysis approach. In the 30 years prior to the opening of the City Loop, the CBD Grid had been steadily losing its share of Melbourne’s employment to other locations. Following construction of the Loop, this downwards trend abated and in the following years, the CBD Grid began to increase its share of employment. The gap between the lowest share and the CBD Grid’s share are used to estimate impact. 3. Transport approach. Levels of rail patronage over time are observed to estimate the capacity of the train network without the City Loop. Every additional job supported beyond the capacity estimate within the CBD Grid can be attributed to the presence of the City Loop.

These methods are conservative, as they assume only development within the CBD can be attributed to the City Loop. Other factors, such as previous planning policies and initiatives such as Postcode 3000 would undoubtedly have boosted employment and housing in the CBD. Box Hill Activities Area For investments into the Box Hill Activities Area, changes to employment have been estimated through comparison with other centres in Melbourne, specifically, Preston in the municipality of Darebin. This has approach helps to conceptualise what Box Hill may have looked and operated like, if it had not received government investment over time. Appendix 1 provides the context for this comparison between Box Hill and Preston.

The accessibility model is then used to estimate the economic impact of investments into Box Hill, with the change in EJD estimated using base travel time matrices and share of public transport use, with altered employment levels in some SLAs. The altered employment is estimated by applying the difference in the level of employment between Box Hill and Preston to the Box Hill SLA. This gives a proxy for the potential outcome in Box Hill had government intervention not occurred. To maintain a fixed Melbourne total of employment, it was assumed that extra jobs removed from Box Hill would be evenly distributed to the surrounding Eastern Melbourne SLAs.

Changes to labour productivity An agglomeration model was used to estimate the economic impact of the CityLink and Western Ring Road projects, using the same estimates of EJD that were used in accessibility modelling. The model uses a Travel Time Matrix, employment and population at a small area level, labour productivity by sector at a State and City level, and then at a small area level. A regression analysis using the EJD estimates and labour productivity by sector is conducted to estimate industry elasticities which show the impact on labour productivity of a doubling in EJD. The change in EJD brought about by a specific infrastructure or land use project can then be established through estimating a project case EJD in the same way the base case EJD is estimated. The inputs into this estimation will differ in terms of a changed travel time matrix or employment numbers.

Finally the impact on gross value added (GVA) arising from changes to labour productivity can be quantified. This is completed via the use of the industry elasticities applied to the estimated changes to EJD to determine the changes to labour productivity. This altered labour productivity then flows through to a change in gross value added across industries and SLAs. Figure 28 shows the process of estimating agglomeration, and Appendix 2 provides further technical notes on this approach.

Confidential Copy 49

FIGURE 28. AGGLOMERATION METHOD – LABOUR PRODUCTIVITY

Labour Productivity by Sector (State and City Level)

Employment by Labour Productivity Travel Time Matrix Sector & Population by Sector (Small Area Level) (Small Area Level) (Small Area)

Calculate Effective Job Density (EJD) (Small Area Level)

Estimate Productivity Elasticity Versus EJD by Sector

Change in EJD Associated with Infrastructure or Land Use Initiative

Estimated Net Increase in GVA

Source: SGS Economics & Planning

There are two components of this benefit to the economy arising for two separate reasons. The first benefit comes from improvements in the relative accessibility of particular locations which have been impacted by the project. This creates improvements in labour productivity in these more accessible areas causing an increase to gross value added (GVA) across metropolitan Melbourne. The second benefit comes from having an increased number of jobs in these more accessible locations that are in more highly productive industries. This will further increase the benefit to GVA. These benefits are additional to the existing GVA of Melbourne and are not simply a redistribution of economic activity within the metropolis. That is, the economy is larger due to the project.

Human Capital Enhancements Firms gain a benefit (which flows onto the rest of the economy) from locating in strategic locations with high levels of accessibility. As firms and households are both economic agents providing services to other economic agents it would stand to reason that households would also obtain a benefit from locating in a strategic location.

The higher metropolitan level data on human capital stock must first be estimated and then broken down to a small area level. This small area level human capital stock data by age, sex and qualification level is combined with EJD in a regression analysis. Elasticities can then be estimated to determine the effect of a change in EJD on human capital stock. The net increase in human capital is then estimated using these three inputs.

Confidential Copy 50

FIGURE 29. AGGLOMERATION METHOD – HUMAN CAPITAL

Human Capital Employment by Stock by Age & Travel Time Matrix Sector & Population Qualification (Small Area Level) (Small Area Level) (Small Area Level)

Calculate Effective Job Density (EJD) (Small Area Level)

Estimate Human Capital Elasticity Versus EJD by Age & Qualification Group

Change in EJD Associated with Infrastructure or Land Use Initiative

Estimated Net Increase in Human Capital

Source: SGS Economics & Planning

The actual productivity growth which will flow from the increased stock of Human Capital is only a proportion of the total estimated uplift. The ABS (2008) estimated that in the most recent (complete) productivity growth cycle6 1998-99 to 2003-04 those improvements to the quality of labour inputs contributed 9.7% of the contribution to growth in real GDP. This national figure of 9.7% has been treated as the ‘elasticity’ for the uplift in Human Capital flowing into the increase in real Melbourne GDP. The final Human Capital benefit is therefore calculated as the estimated uplift in Human Capital multiplied by the population in each education level / age / sex grouping under the project case multiplied by 9.7%. Appendix 2 provides further technical detail on this approach.

Currently there is conjecture as to whether human capital is an additional and separate benefit to the labour productivity improvements already estimated. In this report the benefits have been presented separately as further research is undertaken to provide further clarity around this issue.

Value Chain Accessibility Benefits The increased accessibility and reduced travel times provided by projects have an additional impact to the economy through supply chain improvements. These arise through a reduction in travel times and consequently the cost of freight movements. The input to production cost of road transport will have fallen allowing firms to produce the same level of output at a lower cost. Assuming firms are profit maximising they will produce a higher level of output with the same fixed cost, resulting in an increase in labour productivity. This labour productivity improvement will flow through to an increase in total GVA for Melbourne. This is an additional benefit to those already estimated and to Melbourne’s base case level of economic activity.

This impact has been measured via the use of input-output tables produced by the ABS combined with estimates of GVA, hours worked and labour productivity from the ABS National and State Accounts. The base and project case travel times are the same as those used in the accessibility and agglomeration modelling estimated using the transport modelling program Route Finder. Freight movement data for

6 The long-term trend estimates are calculated using an 11-term Henderson moving average of the original, annual indexes

Confidential Copy 51

2011 was sourced from the Department of Transport (DOT) which provides estimates of the daily truck movements by SLA in an origin destination matrix. To determine the cost of the reduced travel times for road freight an operating cost of $2.107 per kilometre was assumed. The travel times were multiplied by the number of freight trips and the operating cost for each SLA (assuming an average speed of 60km/hour) to estimate the total cost of freight movements. This was completed for the base case (without the project) and project case (with the project) and the change in total cost estimated.

Dwelling and Land Value Impacts Impacts to dwelling and land values can be seen as a proxy for improvements to amenity and accessibility. Where relevant, these impacts have been presented.

To quantify the value of this benefit to the economy SGS has developed an approach using regression analysis. The statistical relationship between house prices (used as a proxy for land values) and effective job density (EJD) was estimated. It was found there was a positive relationship between these two variables. This indicates that in areas that are effectively more dense (higher EJD) in terms of employment or access to employment, house prices will be higher.

Whilst house prices are affected by a variety of factors, this method has controlled for the differences between house prices in the west and south east of Melbourne. A separate regression was run for house and unit prices. This type of analysis only examines the relationship between EJD and house prices and does not attempt to model any of the macroeconomic factors that affect house prices.

In order to estimate the uplift to house and unit prices the regression coefficients were applied to the percentage change estimated for each project. This determined a percentage uplift in median house and unit prices arising from the change in EJD. This percentage change was then applied to the base estimates of median house and unit prices by suburb to determine a project case median house price. The total benefit to Melbourne was then estimated as the sum of the suburb median house price change multiplied by the number of dwellings in each suburb. Given the large variation in house prices the changes to median house prices were capped at the 95th percentile.

Walk time impacts A broader analysis was undertaken to understand the ‘on the ground’ impacts of the City Loop. Walk times were calculated from key City Loop stations along with Southern Cross, Flinders Street, North Melbourne and Jolimont. The analysis shows the dramatic escalation in walking time in the absence of the City Loop, and provides further evidence on how accessibility was improved through the project, shaping the CBD.

7 This estimate was sourced from www.freightmetrics.com.au

Confidential Copy 52

5 IMPACT ASSESSMENT

Drawing from the consolidated dataset and key technical approaches, this section presents analysis on the relationship between each infrastructure project with land use and economic outcomes. The impact of these projects has been assessed via travel time savings, land use changes, as well as the culmination of the impacts into effective job density and industry specific value chain accessibility.

5.1 CityLink

Impacts to Employment and Housing The impacts to employment and housing from CityLink have been estimated using an accessibility model. A map of the changes to effective job density brought about by the CityLink project is presented in Figure 30. This illustrates substantial uplift in EJD in SLAs occurring along the CityLink corridor to the north and to the south east of the CBD.

FIGURE 30. CITYLINK IMPACT TO EFFECTIVE JOB DENSITY, 2011

Source: SGS Economics & Planning

Confidential Copy 53

The impact of this project to the level of employment and households in 2006 and 2011 is presented in Table 5. It is estimated that this project allowed an additional 63,300 jobs in 2006. By 2011 the additional number of jobs increased to 70,300 (that is, an additional 7,000 jobs were added). This was equivalent to 3.2% of the total employment in 2011. There was a larger impact to households and population as a result of this project. It was estimated that approximately 58,200 new households to Melbourne came as a result of CityLink across metropolitan Melbourne in 2011. Figure 31 presents the changes to employment by industry resulting from this project for 2006 and 2011. The largest impacts were estimated to be in Professional, Scientific & Technical Services, Financial & Insurance Services and Health Care & Social Assistance.

TABLE 5. CITYLINK IMPACTS TO EMPLOYMENT AND HOUSEHOLDS Year Employment Households

2006 63,300 52,800 Growth 3.3% 3.7% 2011 70,300 58,200 Growth 3.2% 3.7% Source: SGS Economics & Planning

FIGURE 31. EMPLOYMENT IMPACTS, CITYLINK

14,000

12,000

10,000

8,000

6,000

4,000

2,000

0

2006 2011

Source: SGS Economics & Planning

The distribution of the additional employment impacts are presented in Figure 31 and Table 6 for selected SLAs. The largest increases in employment occurred in Stonnington – Prahran, Moreland – North, Boroondara – Hawthorn and Moonee Valley – West. Most of the SLAs which contain the CityLink roads experienced a positive uplift to employment. This is due to their now increased accessibility to the CBD and many more SLAs compared to before the project was built.

The SLA distribution of the population impacts are presented in Figure 36 and in Table 7. A similar pattern to the distribution of employment impacts can be seen in the distribution of population (and households). Stonnington – Prahran and Boroondara – Hawthorn are estimated to have received an additional 34,500 people in 2011 as a result of the CityLink project. The CityLink project also spurred population growth within the City of Melbourne, of approximately 15,700 people in 2011.

Confidential Copy 54

FIGURE 32. CITYLINK EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

The following there maps show the distribution of the employment impacts in 2011 by broad industry categories. Figure 33 shows the increase in Transport and Wholesale trade industry employment along the CityLink corridor and to the north of the CBD. Figure 34 illustrates the additional knowledge intensive employment brought about by the CityLink project, which includes the industries of Financial & insurance services, Professional, scientific & technical services and Information, media & telecommunications. The impact to population serving industries is presented in Figure 35, which includes the industries of Retail trade, Accommodation & food services, Education & training, Health care & social assistance, Arts & recreation services and Other services.

Confidential Copy 55

FIGURE 33. CITYLINK TRANSPORT & WHOLESALE TRADE EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

ExampleExample of Business ResponseResponse to InfrastructureInfrastructure FinneyFinney & Associates is located in Moonee Ponds and pprovidesrovides accountinaccountingg and financial services to households and businesses. Prior to the openingopening of CityLink,CityLink, FFinneyinney & Associates onlyonly pprovidedrovided services to ppeopleeople livingliving localllocallyy in Moonee ValleValley.y. With the oopeningpening of CitCityLinkyLink FinneyFinney & Associates bebegangan to attract clients from further north and closer to the city.city. The imimprovedproved travel ttimesimes pprovidedrovided by CitCityLinkyLink meant tthathat ppeopleeople coucouldld quicklyquickly travetravell to FFinneyinney & AssociAssociatesates. IncreasinIncreasinglygly more businessesbusinesses were movingmoving into thethe area and were turningturning to FinneyFinney & Associates for advice. In resresponseponse to this increased demand FinneyFinney & Associates took on more staff to deal withwith tthehe personalpersonal taxation demand bbutut aalsolso recruited a bbusinessusiness sspecialistpecialist to take advantaadvantagege of thethe increased opportunitiesopportunities in that area. Profits and wawagesges increased as a result of this increased demand (economies(economies of scale)scale) and the premiumpremium from sspecializationpecialization ((economieseconomies of scoscope)pe) which was a result of the increasedincreased acceaccessibilityssibility pprovidedrovided by CitCityLink.yLink.

Confidential Copy 56

FIGURE 34. CITYLINK KNOWLEDGE INTENSIVE EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

Confidential Copy 57

FIGURE 35. CITYLINK POPULATION SERVING EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

Confidential Copy 58

FIGURE 36. CITYLINK HOUSEHOLD IMPACTS, 2011

Source: SGS Economics & Planning

Confidential Copy 59

TABLE 6. CITYLINK EMPLOYMENT IMPACTS BY SLA, 2006 AND 2011 SLA 2006 Additional Jobs 2006 Growth 2011 Additional Jobs 2011 Growth

Melbourne - Inner 3,600 1.8% 4,600 1.9% Melbourne - S'bank-D'lands 3,400 6.4% 4,400 5.4% Melbourne - Remainder 4,000 3.2% 5,100 3.1% Stonnington - Prahran 18,800 69.6% 19,700 65.7% Yarra - North 1,100 2.8% 1,200 2.9% Yarra - Richmond 900 2.9% 1,000 3.0% Brimbank - Keilor 1,500 6.5% 1,600 6.2% Hobsons Bay - Williamstown 1,800 14.8% 1,900 14.0% Moonee Valley - Essendon 4,300 16.9% 4,500 16.1% Moonee Valley - West 3,700 30.2% 3,900 28.9% Wyndham - South 600 18.7% 600 14.7% Wyndham - West 600 21.8% 600 16.6% Moreland - North 3,200 42.3% 3,400 39.6% Hume - Broadmeadows 1,800 4.0% 1,900 4.2% Hume - Craigieburn 1,700 6.0% 1,800 5.8% Hume - Sunbury 700 9.3% 700 8.6% Whittlesea - North 800 11.4% 800 10.2% Whittlesea - South-East 1,400 19.9% 1,500 18.2% Boroondara - Camberwell S. 900 5.3% 1,000 5.1% Boroondara - Hawthorn 8,800 31.0% 9,300 28.7% Boroondara - Kew 500 3.4% 500 3.1% Stonnington - Malvern 2,500 9.1% 2,600 8.4% Source: SGS Economics & Planning

TABLE 7. CITYLINK POPULATION IMPACTS BY SLA, 2006 AND 2011 SLA 2006 Additional People 2006 Growth 2011 Additional People 2011 Growth

Melbourne - Inner 3,500 28.5% 4,400 27.9% Melbourne - S'bank-D'lands 4,100 29.1% 5,500 28.5% Melbourne - Remainder 5,700 10.5% 5,800 9.1% Stonnington - Prahran 19,400 40.3% 21,500 41.6% Yarra - North 2,600 5.4% 2,900 5.7% Yarra - Richmond 2,100 8.0% 2,300 8.1% Brimbank - Keilor 3,500 3.9% 3,800 4.1% Hobsons Bay - Williamstown 2,700 8.7% 3,000 9.3% Moonee Valley - Essendon 6,300 9.7% 7,100 10.1% Moonee Valley - West 5,600 13.3% 6,100 13.8% Wyndham - South 900 5.5% 1,100 3.3% Wyndham - West 1,000 4.6% 1,100 3.7% Moreland - North 4,700 9.7% 5,100 9.8% Hume - Broadmeadows 4,100 6.3% 4,400 6.7% Hume - Craigieburn 4,100 7.4% 4,500 6.2% Hume - Sunbury 1,400 4.1% 1,500 4.1% Whittlesea - North 1,400 6.0% 1,700 3.1% Whittlesea - South-East 2,700 6.2% 2,900 6.5% Boroondara - Camberwell S. 2,200 4.3% 2,400 4.5% Boroondara - Hawthorn 11,800 33.6% 13,000 34.5% Boroondara - Kew 1,400 4.5% 1,500 4.6% Stonnington - Malvern 4,300 9.1% 4,700 9.6% Source: SGS Economics & Planning

Confidential Copy 60

Economic Expansion Labour Productivity Benefits There are two components of this benefit to the economy arising for two separate reasons. The first benefit comes from improvements in the relative accessibility of particular locations which have been impacted by the project. This creates improvements in labour productivity in these more accessible areas causing an increase to gross value added (GVA) across metropolitan Melbourne. The second benefit comes from having an increased number of jobs in these more accessible locations that are in more highly productive industries. This will further increase the benefit to GVA. Table 8 presents the total benefit which is additional to the existing level of economic activity in Melbourne, as well as the breakdown into the two types of benefits.

When the labour productivity and EJD elasticities are applied to the project induced changes in EJD, the total impact on metropolitan GVA was estimated to be $7.89 billion in 2006 and $8.95 billion in 2011. Of this $8.95 billion benefit in 2011, 16% ($1.39 billion) arises from improvements to labour productivity, with the remaining 84% ($7.55 billion) coming from the additional employment benefit.

It should be noted that the $1.4 billion Labour Productivity Benefit would be split very roughly 50-50 between high profits and higher wages. Working off an average tax rate of 32%8 would result in an additional tax receipt for the Commonwealth of $448 million. The same cannot be said with certainty for the additional employment benefit. It is possible that some (or all) of these jobs may have appeared elsewhere in Australia. And increased Commonwealth taxes can’t be attributed to the income being generated by these jobs.

TABLE 8. CITYLINK IMPACTS TO MELBOURNE GROSS VALUE ADDED ($000’S) 2006 2011

Labour Productivity Benefit 1,149,390 1,395,310 Share of Total Benefit 15% 16% Additional Employment Benefit 6,735,980 7,557,450 Share of Total Benefit 85% 84% Total Benefit 7,885,370 8,952,760 Source: SGS Economics & Planning

To examine the robustness of the additional employment benefits a sensitivity analysis was conducted. This analysis varied the amount of the additional employment that was new to Melbourne. The $8.95 billion benefit (in 2011) referred to above and in Table 8 assumes that the additional 70,300 jobs are new to Melbourne. An alternate scenario assumes that only 50% of these jobs will be additional to Melbourne. The impact to GVA under this scenario was estimated to be $3.77 billion in 2011.

A third scenario assumes that none of these jobs will be additional. It instead assumes they have been redistributed throughout Melbourne with the total employment fixed at base case levels. The impact to GVA under this scenario was estimated to be negative due to the new distribution of employment. Compared to the base case this redistribution is less efficient and places more employment in less productive areas thus causing the total output to be lower.

8 This share requires further cross checks.

Confidential Copy 61

TABLE 9. SENSITIVITY ANALYSIS OF ADDITIONAL EMPLOYMENT BENEFIT, CITYLINK ($000’S) 2006 2011

100% Additional Employment 6,735,980 7,546,920 50% Additional Employment 3,366,570 3,774,340 0% Additional Employment -21,390 -73,660 Source: SGS Economics & Planning

Table 10 presents the benefits by industry in 2006 and 2011 for all of metropolitan Melbourne under the 100% additional employment scenario. The largest benefit is forecast to flow from Financial & Insurance Services ($2.2 billion in 2011) and Professional, Scientific & Technical services ($1.51 billion). The project has small adverse impacts on the Agriculture, Forestry & Fishing industry. This is consistent with historical trends which have seen these types of space intensive industries displaced from more central locations (where EJD is high) to less central locations (where EJD is lower).

TABLE 10. CITYLINK IMPACTS ON INDUSTRY GROSS VALUE ADDED ($000’S), MELBOURNE Industry 2006 Benefit 2006 Share of Benefit 2011 Benefit 2011 Share of Benefit

Agriculture, Forestry & Fishing -41,260 -0.5% -36,250 -0.4% Mining 43,350 0.5% 41,970 0.5% Manufacturing 227,380 2.9% 197,470 2.2% Electricity, Gas, Water & Waste 193,740 2.5% 321,050 3.6% Construction 352,600 4.5% 386,470 4.3% Wholesale Trade 116,630 1.5% 127,780 1.4% Retail Trade 350,300 4.4% 367,090 4.1% Accommodation & Food 232,180 2.9% 282,530 3.2% Transport, Postal & Warehousing 279,320 3.5% 333,870 3.7% Information Media & Telecom. 804,890 10.2% 807,130 9.0% Financial & Insurance 1,926,650 24.4% 2,213,060 24.7% Rental, Hiring & Real Estate 299,990 3.8% 317,240 3.5% Professional, Scientific & Technical 1,277,700 16.2% 1,514,910 16.9% Administrative & Support 453,770 5.8% 420,150 4.7% Public Admin & Safety 343,090 4.4% 457,190 5.1% Education & Training 253,650 3.2% 323,770 3.6% Health Care & Social Assist. 432,840 5.5% 524,010 5.9% Arts & Recreation 213,770 2.7% 216,290 2.4% Other Services 124,780 1.6% 137,050 1.5% Total 7,885,370 100% 8,952,780 100% Source: SGS Economics & Planning

In absolute terms, the boost to SLA value added for the project under the 100% scenario would have been strongest in the three Melbourne SLAs, Stonnington – Prahran and Boroondara – Hawthorn. In proportion to SLA size terms, the biggest impacts in GVA are expected to have been felt in Stonnington – Prahran (47.9% increase in SLA GVA in 2011), Moreland – North (34.6%) and Boroondara – Hawthorn (25.9%). These GVA impacts by SLA are presented in Table 11, Figure 37 and Figure 38.

Confidential Copy 62

FIGURE 37. CITYLINK AGGLOMERATION BENEFITS, 2011

Source: SGS Economics & Planning

Confidential Copy 63

FIGURE 38. CITYLINK ADDITIONAL EMPLOYMENT BENEFITS, 2011

Source: SGS Economics & Planning

Confidential Copy 64

TABLE 11. CITYLINK GROSS VALUE ADDED IMPACTS BY SLA, 2006 AND 2011 ($000’S) SLA 2006 Benefit 2006 Share of GVA 2011 Benefit 2011 Share of GVA

Melbourne - Inner 707,120 2.3% 869,340 2.5% Melbourne - S'bank-D'lands 548,280 8.5% 743,930 5.5% Melbourne - Remainder 731,390 5.8% 911,770 5.4% Stonnington - Prahran 2,164,310 87.1% 2,304,490 47.9% Yarra - North 135,720 4.0% 151,160 4.0% Yarra - Richmond 88,360 2.6% 103,960 2.7% Brimbank - Keilor 145,680 8.7% 156,160 7.8% Hobsons Bay - Williamstown 182,910 18.0% 195,110 15.0% Moonee Valley - Essendon 431,470 23.1% 461,380 18.5% Moonee Valley - West 323,340 39.7% 342,840 28.1% Wyndham - South 50,090 23.1% 51,470 16.8% Wyndham - West 41,920 20.8% 44,160 15.2% Moreland - North 259,980 54.7% 274,160 34.6% Hume - Broadmeadows 171,630 4.8% 188,530 4.9% Hume - Craigieburn 181,250 7.5% 192,240 6.9% Hume - Sunbury 58,310 13.4% 62,260 11.2% Nillumbik - South-West 38,940 13.4% 41,180 11.0% Nillumbik Bal 19,820 14.4% 21,410 12.0% Whittlesea - North 74,690 11.4% 76,150 9.5% Whittlesea - South-East 110,960 25.7% 117,850 19.3% Boroondara - Camberwell S. 97,350 6.9% 105,710 6.2% Boroondara - Hawthorn 1,058,660 35.9% 1,124,520 25.9% Boroondara - Kew 56,260 4.5% 60,990 4.0% Stonnington - Malvern 271,010 13.2% 292,330 11.1% Source: SGS Economics & Planning

Human Capital Enhancements

Table 12 presents the total human capital impacts flowing from the CityLink project. This represents the benefit to all of metropolitan Melbourne, estimated at $12.5 million in 2006 and $13.9 million in 2011. This total benefit to human capital is made up of benefits flowing to different levels of qualification groups, with the largest being from Bachelor Degree Males ($3.4 million in 2011). Figure 39 presents a map of the benefits by SLA across Melbourne.

These findings regarding the link between improved jobs accessibility enabled by the project and enrichment of the city’s human capital stock resonates with other SGS investigations of these issues. By relating house prices to jobs accessibility, SGS has previously found that, on average, Melbourne households (outside of the inner urban core) are willing to pay around 43 cents for each additional job brought within a 30 minute drive time. This evident willingness to pay for greater jobs choice reflects, at least in part, confidence on the part of the households in question that they will enjoy superior permanent income prospects when located in jobs rich or highly accessible districts.

Confidential Copy 65

TABLE 12. CITYLINK HUMAN CAPITAL IMPACTS TO GVA, 2006 AND 2011 Qualification Level 2006 2011

Male Unqualified $1,179,700 $1,329,200 Skilled Labour $1,058,800 $1,200,300 Bachelor Degree $3,148,600 $3,488,600 Higher Degree $868,600 $959,900 Female Unqualified $1,791,100 $2,018,500 Skilled Labour $1,066,200 $1,197,700 Bachelor Degree $2,874,200 $3,189,100 Higher Degree $526,600 $582,700 Total Unqualified $2,970,800 $3,347,700 Skilled Labour $2,125,000 $2,398,000 Bachelor Degree $6,022,800 $6,677,700 Higher Degree $1,395,200 $1,542,600 Total Benefit $12,513,800 $13,966,000 Source: SGS Economics & Planning

FIGURE 39. CITYLINK HUMAN CAPITAL BENEFITS, 2011

Source: SGS Economics & Planning

Confidential Copy 66

Value Chain Accessibility Benefits The increased accessibility and reduced travel times provided by the CityLink project have an additional impact to the economy through supply chain improvements. These arise through a reduction in travel times and consequently the cost of freight movements. The input to production cost of road transport will have fallen allowing firms to produce the same level of output at a lower cost. Assuming firms are profit maximising they will produce a higher level of output with the same fixed cost, resulting in an increase in labour productivity. This labour productivity improvement will flow through to an increase in total gross value added (GVA) for Melbourne.

This percentage change was assumed to be the reduction in input costs of road transport for all other industries. For the CityLink project it was estimated to be a 0.8% reduction in road transport costs. This flows through to a small increase in labour productivity across all industries, resulting in a higher level of GVA in particular industries. Following this GVA increase employment will also increase as certain industries become more attractive to workers. Gross value added will be increased further by this change in employment and hours worked. The total reduction in travel times as a result of the project across the entire network was estimated to be 0.84%. This was largely concentrated in the SLAs of Melbourne, Stonnington, Monash, Brimbank, Moonee Valley and Hume.

The total impacts to gross value added from the improved value chain accessibility are presented in Table 24 for 2006 and 2011 for all of metropolitan Melbourne. In 2011 the impact of the project was estimated to be $80.6 million, equivalent to 0.04% of GVA in Melbourne. This benefit can be broken down into industry benefits which vary based on their dependence on road transport as an input. The industries which gain the largest benefit include Manufacturing (0.24% of GVA impact), Construction (0.05%), Accommodation & Food Services (0.05%) and Transport, Postal & Warehousing (0.04%). Table 14 presents these benefits by industry for 2006 and 2011 for metropolitan Melbourne.

TABLE 13. CITYLINK VALUE CHAIN ACCESSIBILITY BENEFITS ($MILLION) 2006 2011

Total Benefit 76.7 80.6 Share of GVA 0.04% 0.04% Source: SGS Economics & Planning

Confidential Copy 67

TABLE 14. CITYLINK VALUE CHAIN ACCESSIBILITY BENEFITS TO INDUSTRY GROSS VALUE ADDED ($MILLION) Industry 2006 Benefit 2006 Share of GVA 2011 Benefit 2011 Share of GVA

Agriculture, Forestry & Fishing 0.5 0.05% 0.4 0.05% Mining 0.1 0.02% 0.1 0.02% Manufacturing 49.1 0.24% 49.0 0.24% Electricity, Gas, Water & Waste 1.5 0.04% 1.6 0.04% Construction 6.5 0.05% 7.8 0.05% Wholesale Trade 2.1 0.02% 2.3 0.02% Retail Trade 1.4 0.01% 1.6 0.01% Accommodation & Food 2.4 0.05% 2.7 0.05% Transport, Postal & Warehousing 4.2 0.04% 4.6 0.04% Information Media & Telecom. 1.7 0.02% 2.0 0.02% Financial & Insurance 0.2 0.00% 0.3 0.00% Rental, Hiring & Real Estate 0.3 0.01% 0.3 0.01% Professional, Scientific & Technical 1.4 0.01% 2.0 0.01% Administrative & Support 0.6 0.01% 0.7 0.01% Public Admin & Safety 1.1 0.01% 1.2 0.01% Education & Training 1.3 0.01% 1.5 0.01% Health Care & Social Assist. 0.8 0.01% 1.0 0.01% Arts & Recreation 0.9 0.04% 1.1 0.04% Other Services 0.5 0.01% 0.5 0.01% Total 76.7 0.04% 80.6 0.04% Source: SGS Economics & Planning

Increase in Land Value The average uplift in median house prices across Melbourne was estimated to be $25,400 as a result of the project. Using estimates from the housing development data (HDD) survey conducted in 2009, the total number of houses across Melbourne was 1.22 million. This translated to a once off total benefit to the economy of $19.3 billion.

These estimates are presented in Table 15 along with the results for units. The total uplift in house prices in the inner region was estimated to be $284 million, calculated as the number of dwellings in this region (3,500) multiplied by the average median house price uplift. This was done at the suburb level as each suburb in the inner region had median house prices varying around the average of $66,300.

Figure 40 and Figure 41 present the spatial distribution of these land value impacts across Melbourne showing the suburbs most affected by the CityLink project. The largest impacts flow from the suburbs located along the CityLink corridor and to the north, whilst smaller impacts are more widespread. As shown in Table 15 the western, eastern and northern regions experience the largest impact to house prices. Due to the different housing types across Melbourne the inner region contains a larger number of units than houses. This causes the total uplift for units to be larger in the inner and eastern regions.

Confidential Copy 68

TABLE 15. CITYLINK LAND VALUE IMPACTS Housing Region Average Uplift to Median House Number of Dwellings Total Uplift ($) Type Prices ($)

Inner 66,300 3,500 284,456,200 Eastern 16,300 363,200 6,373,894,000 Western 36,000 219,500 6,600,486,100 Southern 2,100 391,900 555,518,900 House House Northern 22,900 242,000 5,475,322,600 Total Melbourne 25,400 1,220,100 19,289,677,800 Inner 57,900 59,900 3,120,171,400 Eastern 15,200 81,500 3,583,865,100 Western 26,200 37,500 1,773,584,600

Unit Southern 1,600 88,200 905,379,300 Northern 23,900 43,200 973,520,900 Total Melbourne 18,400 310,300 10,356,521,300 Source: SGS Economics & Planning

FIGURE 40. CITYLINK IMPACTS TO HOUSE PRICES, 2011

Source: SGS Economics & Planning

Based on average rate receipts for local government from land taxes this increased value of properties could add around $300 million dollars to the coffers of local governments every year, if they set their

Confidential Copy 69

outlays growth in line with property value increases. It would also add to higher stamp duties for the State government.

FIGURE 41. CITYLINK IMPACTS TO UNIT PRICES, 2011

Source: SGS Economics & Planning

Example of Business Response to InfrastructureInfrastructure DuringDuring tthehe llateate 1990s Jovanovic ConstructioConstructionn sspecialisedpecialised in devedevelopingloping smasmallll scascalele susubdivisionsbdivisions in inner eastern Melbourne. Due to a shortashortagege of suitable ssitesites ththee ffeasibilityeasibility of the construction in ththee inner eastern suburbs was becomingbecoming increasinglyincreasingly difficult. However, the openingopening of CitCityLinkyLink oopenedpened uupp a ranrangege of small scale developmentdevelopment oopportunitiespportunities in CitCityy of StoninStoningtongton and surrounds. TThishis wwasas due toto thethe improvedimproved accessiaccessibilitybility to tthehe ememploymentployment oopportunitiespportunities in tthehe citcityy and further to the south east increasinincreasingg the demand for ppeopleeople wishinwishingg to locate close to CityLink.CityLink. That is, there was an increase in underlunderlyingying land values in these locations. The firm increasedincreased ooperationsperations and alsoalso saw thethe opportunitiesopportunities to undertakeundertake smsmaallll scascalele devedevelopmentslopments in tthehe NNorthorth WestWest corridor of CitCityLink.yLink. This was a ppartart of the citycity which did not have a historyhistory of medium densitydensity development.development. Jovanovic ConstructioConstructionn qquicklyuickly established itself as a niche pplayerlayer in the propertyproperty market and rarapidlypidly exexpanded.panded. SeeinSeeingg that there was considerable aappetiteppetite for medium densitydensity llevelevel JovanovicJovanovic CConstructiononstruction moved into largelarge scalescale developmentsdevelopments constructingconstructing five storystory apartmentapartment buildingsbuildings in a ranrangege of locations in Melbourne.

Confidential Copy 70

5.2 Western Ring Road

Impacts to Employment and Housing Figure 42 presents the estimated changes to effective job density for the Western Ring Road project. This illustrates the larger benefit the north and west of Melbourne receives than compared to the south eastern region.

FIGURE 42. WESTERN RING ROAD IMPACTS TO EFFECTIVE JOB DENSITY, 2011

Source: SGS Economics & Planning

The impact of this project to the level of employment and households in 2006 and 2011 is presented in Table 16. It is estimated that this project allowed an additional 24,900 jobs in 2006 and 27,300 jobs to occur in Melbourne in 2011. This was equivalent to 1.3% of the total employment in each year. There was a similar impact to population and households as a result of this project. It was estimated that approximately 19,300 additional households moved to Melbourne as a result of the Western Ring Road in 2011. Figure 43 presents the changes to employment by industry resulting from this project for 2006 and 2011. The largest impacts were estimated to be in Professional, Scientific & Technical Services, Financial & Insurance Services and Health Care & Social Assistance.

Confidential Copy 71

TABLE 16. WESTERN RING ROAD IMPACTS TO EMPLOYMENT AND HOUSEHOLDS Year Employment Population

2006 24,900 17,500 Growth 1.3% 1.2% 2011 27,300 19,300 Growth 1.3% 1.2% Source: SGS Economics & Planning

FIGURE 43. EMPLOYMENT IMPACTS, WESTERN RING ROAD

6,000

5,000

4,000

3,000

2,000

1,000

0

2006 2011

Source: SGS Economics & Planning

The distribution of the additional employment impacts are presented in Figure 44 and Table 17 for selected SLAs. The largest increases in employment occurred in Hume – Craigieburn, Melbourne – Remainder and Brimbank – Sunshine. Most of the SLAs which contain on and off ramps to the Western Ring Road experienced a positive uplift to employment. This is due to their now increased accessibility from the CBD to the West compared to before the project was built.

The SLA distribution of the population impacts are presented in the map in Figure 48 and Table 18. The majority of the impact to population was felt in the Western suburbs, particularly in Hume – Craigieburn, Whittlesea – South East, Brimbank – Sunshine and Mooney Valley – West. These four SLAs are estimated to have allowed a total of 16,500 additional people (28% of the total impact) in 2006 as a result of the project.

Confidential Copy 72

FIGURE 44. WESTERN RING ROAD EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

The following there maps show the distribution of the employment impacts in 2011 by broad industry categories. Figure 45 shows the increase in Transport and Wholesale trade industry employment along the Western Ring Road corridor and to the north and west of Melbourne. Figure 34 illustrates the additional knowledge intensive employment brought about by the Western Ring Road project, which includes the industries of Financial & insurance services, Professional, scientific & technical services and Information, media & telecommunications. The impact to population serving industries is presented in Figure 47, which includes the industries of Retail trade, Accommodation & food services, Education & training, Health care & social assistance, Arts & recreation services and Other services.

Confidential Copy 73

FIGURE 45. WESTERN RING ROAD TRANSPORT & WHOLESALE TRADE EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

Example of Business ResponseResponse to InfrastructureInfrastructure In tthehe mid-1990smid-1990s TEX InIndustrialdustrial waswas llocatedocated in WestWest MelbourneMelbourne notnot farfar fromfrom ththee PPortort ooff MeMelbourne.lbourne. TEX IndustriaIndustriall operatedoperated out of a 1199th CCenturyentury warewarehousehouse wwhichhich was aalsolso owned by thethe comcompany.pany. The warehouse was poorlypoorly designeddesigned for their ppurposesurposes which lead to inefficient processingprocessing and handlinhandlingg of ggoods.oods. But the need for ggoodood access to the Port meant that the West Melbourne location was ideal for the company.company. With the construction and openingopening of the WesternWestern RingRing RoadRoad TEX IndustrialIndustrial saw an oopportunitypportunity to relocate to a ppurposeurpose built facilitfacilityy in Laverton which still has susuperbperb access to the Port of Melbourne. The new facilitfacilityy was able to boost their productivityproductivity bbyy 10 pperer cent and have a hihighgh level of accessibilitaccessibilityy to firms in western MelbourneMelbourne whilewhile stillstill rretainetain access ttoo ththee PPort.ort. ThThee ddifferenceifference in llandand vvaleale mmeanteant ththatat ththee sasalele ooff thethe site in West Melbourne unlocked enouenoughgh cacapitalpital to cover the construction of the facilitfacilityy in Laverton. FollowingFollowing the departuredeparture of TEX IndustrialIndustrial thethe warehousewarehouse waswas convertedconverted intintoo apartments.apartments.

Confidential Copy 74

FIGURE 46. WESTERN RING ROAD KNOWLEDGE INTENSIVE EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

Confidential Copy 75

FIGURE 47. WESTERN RING ROAD POPULATION SERVING EMPLOYMENT IMPACTS, 2011

Source: SGS Economics & Planning

Confidential Copy 76

FIGURE 48. WESTERN RING ROAD HOUSEHOLD IMPACTS, 2011

Source: SGS Economics & Planning

Confidential Copy 77

TABLE 17. WESTERN RING ROAD EMPLOYMENT IMPACTS BY SLA, 2006 AND 2011 SLA 2006 Additional Jobs 2006 Growth 2011 Additional Jobs 2011 Growth

Melbourne - Remainder 3,300 2.7% 4,000 2.5% Brimbank - Keilor 1,400 6.0% 1,400 5.4% Brimbank - Sunshine 3,200 9.9% 3,400 9.4% Hobsons Bay - Altona 1,100 4.3% 1,100 4.2% Hobsons Bay - Williamstown 800 6.8% 900 6.4% Moonee Valley - West 2,200 18.1% 2,300 17.2% Melton - East 1,000 20.6% 1,000 17.5% Wyndham - South 600 21.1% 700 17.0% Wyndham - West 700 25.3% 700 19.4% Moreland - Coburg 700 5.3% 700 4.8% Moreland - North 2,100 27.8% 2,200 25.6% Banyule - North 600 4.4% 600 3.9% Hume - Broadmeadows 1,400 3.3% 1,500 3.3% Hume - Craigieburn 2,700 9.5% 2,900 9.2% Hume - Sunbury 500 6.1% 500 5.4% Nillumbik - South 300 3.9% 300 3.5% Whittlesea - South-East 2,100 29.6% 2,200 26.8% Whittlesea - South-West 1,300 5.1% 1,300 4.9% Source: SGS Economics & Planning

TABLE 18. WESTERN RING ROAD POPULATION IMPACTS BY SLA, 2006 AND 2011 SLA 2006 Additional People 2006 Growth 2011 Additional People 2011 Growth

Melbourne - Remainder 1,000 1.9% 1,100 1.7% Brimbank - Keilor 2,500 2.7% 2,700 2.9% Brimbank - Sunshine 5,000 5.9% 5,500 5.5% Hobsons Bay - Altona 1,600 3.0% 1,700 3.1% Hobsons Bay - Williamstown 1,100 3.7% 1,300 3.9% Moonee Valley - West 3,100 7.5% 3,400 7.8% Melton - East 1,500 3.8% 1,800 2.9% Wyndham - South 900 5.4% 1,100 3.2% Wyndham - West 1,000 4.5% 1,100 3.6% Moreland - Coburg 1,100 2.1% 1,200 2.2% Moreland - North 2,900 6.0% 3,100 6.0% Banyule - North 1,000 1.8% 1,000 1.8% Hume - Broadmeadows 2,600 4.1% 2,800 4.3% Hume - Craigieburn 4,900 8.8% 5,400 7.4% Hume - Sunbury 700 2.2% 800 2.2% Nillumbik - South 600 2.1% 600 2.2% Whittlesea - South-East 3,500 7.9% 3,700 8.3% Whittlesea - South-West 2,200 3.6% 2,400 3.8% Source: SGS Economics & Planning

Economic Expansion Labour Productivity Benefits Table 19 presents the total benefit additional to Melbourne as well as the breakdown into the two benefits of improvements to labour productivity and additional employment in more accessible locations in more highly productive industries.

When the labour productivity and EJD elasticities are applied to the project induced changes in EJD, the total impact on metropolitan GVA was estimated to be $2.45 billion in 2006 and $2.68 billion in 2011. Of this $2.68 billion benefit in 2011, 92% ($2.45 billion) arises from the additional employment benefits with the remaining 8% coming from improvements to labour productivity.

Confidential Copy 78

It should be noted that the $227.8 million increase in GVA would be split very roughly 50-50 between high profits and higher wages. Working off an average tax rate of 32% this would result in additional tax receipts for the Commonwealth of $72 million.

TABLE 19. WESTERN RING ROAD IMPACTS TO MELBOURNE GROSS VALUE ADDED ($000’S) 2006 2011

Labour Productivity Improvements 191,630 227,760 Share of Total Benefit 8% 8% Additional Employment Benefit 2,266,240 2,457,970 Share of Total Benefit 92% 92% Total Benefit 2,457,870 2,685,730 Source: SGS Economics & Planning

The same sensitivity analysis was undertaken for the Western Ring Road project additional employment benefits as outlined for the CityLink project. This analysis varied the amount of the additional employment that was new to Melbourne. The $2.46 billion benefit (in 2011) referred to above and in Table 19 assumes that the additional 24,900 jobs are new to Melbourne. An alternate scenario assumes that only 50% of these jobs will be additional to Melbourne. The impact to GVA under this scenario was estimated to be $1.23 billion in 2011.

A third scenario assumes that none of these jobs will be additional. It instead assumes they have been redistributed throughout Melbourne with the total employment fixed at base case levels. The impact to GVA under this scenario was estimated to be negative due to the new distribution of employment. Compared to the base case this redistribution is less efficient and places more employment in less productive areas thus causing the total output to be lower. This can be seen as a much less likely scenario as the majority of these jobs appear to be new to Melbourne rather than just redistributed within Melbourne.

TABLE 20. SENSITIVITY ANALYSIS OF ADDITIONAL EMPLOYMENT BENEFIT, WESTERN RING ROAD ($000’S) 2006 2011

100% Additional Employment 2,266,240 2,457,970 50% Additional Employment 1,132,490 1,229,120 0% Additional Employment -404,590 -434,990 Source: SGS Economics & Planning

Table 21 presents the benefits by industry in 2006 and 2011 for all of metropolitan Melbourne. The largest benefit is forecast to flow from Financial & Insurance services ($597.3 million), Professional, Scientific & Technical Services ($356.5 million) and Information, Media & Telecommunications ($266.9 million).

Confidential Copy 79

TABLE 21. WESTERN RING ROAD IMPACTS ON INDUSTRY GROSS VALUE ADDED ($000’S), MELBOURNE Industry 2006 Benefit 2006 Share of Benefit 2011 Benefit 2011 Share of Benefit

Agriculture, Forestry & Fishing -6,030 -0.2% -6,310 -0.2% Mining 15,230 0.6% 12,460 0.5% Manufacturing 69,930 2.8% 91,340 3.4% Electricity, Gas, Water & Waste 64,740 2.6% 97,780 3.6% Construction 106,490 4.3% 125,750 4.7% Wholesale Trade 50,600 2.1% 65,200 2.4% Retail Trade 79,250 3.2% 94,010 3.5% Accommodation & Food 77,930 3.2% 91,680 3.4% Transport, Postal & Warehousing 116,790 4.8% 142,560 5.3% Information Media & Telecom. 287,690 11.7% 266,950 9.9% Financial & Insurance 580,600 23.6% 597,360 22.2% Rental, Hiring & Real Estate 69,400 2.8% 71,660 2.7% Professional, Scientific & Technical 332,760 13.5% 356,500 13.3% Administrative & Support 169,370 6.9% 138,060 5.1% Public Admin & Safety 123,090 5.0% 153,190 5.7% Education & Training 93,860 3.8% 129,220 4.8% Health Care & Social Assist. 125,170 5.1% 157,650 5.9% Arts & Recreation 63,170 2.6% 55,920 2.1% Other Services 37,840 1.5% 44,740 1.7% Total 2,457,880 100% 2,685,720 100% Source: SGS Economics & Planning

In absolute terms of the boost to SLA value added, the project had the strongest impact on Whittlesea – South-East, Wyndham – North, Moreland – North and Hume – Craigieburn. Proportional to SLA size terms, the biggest impacts in GVA were felt in Whittlesea – South East, Wyndham – North and Moreland – North. These GVA impacts by SLA are presented in Table 22, Figure 49 and Figure 50.

Confidential Copy 80

FIGURE 49. WESTERN RING ROAD AGGLOMERATION BENEFITS, 2011

Source: SGS Economics & Planning

Confidential Copy 81

FIGURE 50. WESTERN RING ROAD ADDITIONAL EMPLOYMENT BENEFITS, 2011

Source SGS Economics & Planning

Confidential Copy 82

TABLE 22. WESTERN RING ROAD GROSS VALUE ADDED IMPACTS BY SLA, 2006 AND 2011 SLA 2006 Benefit 2006 Share of GVA 2011 Benefit 2011 Share of GVA

Melbourne - Remainder 519,109 4.0% 605,317 3.6% Brimbank - Keilor 125,842 7.0% 127,889 6.5% Brimbank - Sunshine 293,808 10.1% 305,591 9.6% Hobsons Bay - Altona 96,292 4.5% 100,039 4.5% Hobsons Bay - Williamstown 84,167 7.7% 86,670 7.2% Moonee Valley - West 192,809 19.3% 199,348 18.3% Melton - East 81,791 19.5% 83,075 16.9% Wyndham - South 55,793 6.1% 56,649 5.6% Wyndham - West 48,486 26.6% 49,950 24.9% Moreland - Coburg 63,168 6.0% 63,512 5.3% Moreland - North 170,047 3.7% 173,999 3.7% Banyule - North 50,900 10.2% 51,262 9.8% Hume - Broadmeadows 135,837 7.8% 142,854 7.0% Hume - Craigieburn 270,573 5.1% 280,672 4.5% Hume - Sunbury 36,739 27.7% 37,096 25.4% Nillumbik - South 28,248 5.7% 28,813 5.5% Whittlesea - South-East 162,330 4.0% 167,347 3.6% Whittlesea - South-West 114,390 7.0% 119,846 6.5% Source: SGS Economics & Planning

Human Capital Enhancements Table 23 presents the total human capital impacts flowing from the Western Ring Road project. This represents the benefit to all of metropolitan Melbourne, estimated at $3.9 million in 2006 and $4.6 million in 2011. This total benefit to human capital is made up of benefits flowing to different levels of qualification groups, with the largest being from Unqualified Females ($1,051,000). Figure 51 presents a map of the benefits by SLA across Melbourne.

TABLE 23. WESTERN RING ROAD HUMAN CAPITAL IMPACTS TO GVA, 2006 AND 2011 Qualification Level 2006 2011

Male Unqualified $587,300 $699,000 Skilled Labour $519,700 $628,200 Bachelor Degree $612,400 $717,400 Higher Degree $132,100 $153,600 Female Unqualified $884,300 $1,051,000 Skilled Labour $444,300 $531,300 Bachelor Degree $621,100 $728,600 Higher Degree $78,200 $90,200 Total Unqualified $1,471,600 $1,750,000 Skilled Labour $964,000 $1,159,500 Bachelor Degree $1,233,500 $1,446,000 Higher Degree $210,300 $243,800 Total Benefit $3,879,400 $4,599,300 Source: SGS Economics & Planning

Confidential Copy 83

ExampleExample ofof Business ResponseResponse to InfrastructureInfrastructure MarcusMarcus finished hihighgh school at St Albans SecondarySecondary CollegeCollege in 1999. DuringDuring his time at highhigh scschoolhool hhee hadhad a partpart time jobjob at a locallocal restaurant. BeinBeingg ggoodood witwithh numnumbersbers meant tthathat MMarcusarcus hhadad workedworked out tthathat hhee shouldshould spendspend aroundaround $30$30 a week on petrolpetrol in gettinggetting to work. However, tthehe uncertaintyuncertainty surroundingsurrounding the oncomingoncoming introduction of the GST in June 2000 had made local ememployersployers in and around St Albans reluctant to take on new staff. MMarcusarcus struggledstruggled to find ememploymentployment durinduringg 2000 and keptkept workingworking at hishis partpart time jobjob at thethe restaurant. However, thethe oopeningpening of the Western RingRing Road in 2001 meant that MMarcusarcus widened his search for ememploymentployment and found a jjobob near Melbourne AirAirportport workinworkingg for a firm pprovidingroviding infliinflightght meamealsls ttoo tthehe airlines.airlines.

FIGURE 51. WESTERN RING ROAD HUMAN CAPITAL BENEFITS, 2011

Source: SGS Economics & Planning

Value Chain Accessibility Benefits It is estimated that the Western Ring Road project resulted in a 1.6% reduction in road transport costs, leading to wider benefits through higher labour productivity and GVA. The total reduction in travel times as a result of the project across the entire network was estimated to be 1.55%. This was largely concentrated in the SLAs of Brimbank, Moonee Valley, Wyndham, Moreland and Hume.

Confidential Copy 84

The total impacts to gross value added (GVA) from the improved value chain accessibility are presented in Table 24 for 2006 and 2011 for all of metropolitan Melbourne. In 2011 the impact of the project was estimated to be $149.3 million, equivalent to 0.07% of GVA in Melbourne. This benefit can be broken down into industry benefits which vary based on their dependence on road transport as an input. The industries which gain the largest benefit include Manufacturing (0.44% of GVA impact), Construction (0.10%), Transport, Postal & Warehousing (0.08%) and Accommodation & Food Services (0.09%). Table 25 presents these benefits by industry for 2006 and 2011 for metropolitan Melbourne.

TABLE 24. WESTERN RING ROAD VALUE CHAIN ACCESSIBILITY BENEFITS ($MILLION) 2006 2011

Total Benefit 142.0 149.3 Share of GVA 0.08% 0.07% Source: SGS Economics & Planning

TABLE 25. WESTERN RING ROAD VALUE CHAIN ACCESSIBILITY BENEFITS, INDUSTRY GVA ($MILLION) Industry 2006 Benefit 2006 Share of GVA 2011 Benefit 2011 Share of GVA

Agriculture, Forestry & Fishing 0.9 0.08% 0.8 0.08% Mining 0.1 0.04% 0.2 0.04% Manufacturing 91.0 0.44% 90.8 0.44% Electricity, Gas, Water & Waste 2.8 0.07% 3.0 0.07% Construction 12.1 0.10% 14.5 0.10% Wholesale Trade 4.0 0.04% 4.3 0.04% Retail Trade 2.7 0.03% 2.9 0.03% Accommodation & Food 4.5 0.09% 4.9 0.09% Transport, Postal & Warehousing 7.8 0.08% 8.5 0.08% Information Media & Telecom. 3.1 0.04% 3.7 0.04% Financial & Insurance 0.4 0.00% 0.5 0.00% Rental, Hiring & Real Estate 0.6 0.01% 0.6 0.01% Professional, Scientific & Technical 2.6 0.02% 3.6 0.02% Administrative & Support 1.1 0.02% 1.2 0.02% Public Admin & Safety 2.0 0.03% 2.3 0.03% Education & Training 2.5 0.02% 2.8 0.02% Health Care & Social Assist. 1.4 0.01% 1.8 0.01% Arts & Recreation 1.6 0.07% 2.0 0.07% Other Services 0.9 0.02% 0.9 0.02% Total 142.0 0.08% 149.3 0.07% Source: SGS Economics & Planning

Increase in Land Values The average uplift in median house prices across Melbourne was estimated to be $6,800 as a result of the project. Using estimates from the housing development data (HDD) survey conducted in 2009, the total number of houses across Melbourne was 1.22 million. This translated to a once off total benefit to the economy of $9.18 billion.

These estimates are presented in Table 26 along with the results for units. The total uplift in house prices in the western region was estimated to be $5.09 billion calculated as the number of dwellings in this region (219,500) multiplied by the average median house price uplift. This was done at the suburb level as each suburb in the western region had median house prices varying around the average of $23,000.

Figure 52 and Figure 53 present the spatial distribution of these land value impacts across Melbourne showing the suburbs most affected by the Western Ring Road project. The largest impacts flow from the suburbs located along the highway corridor and to the west, whilst smaller impacts are more widespread to the north. As shown in Table 26 the western and northern regions experienced the largest impact to

Confidential Copy 85

house prices. The inner, eastern and southern regions were not affected by this project in terms of land value impacts.

TABLE 26. WESTERN RING ROAD LAND VALUE IMPACTS Housing Region Average Uplift to Median Number of Dwellings Total Uplift Type House Prices

Inner 0 3,500 0 Eastern 0 363,200 0 Western 23,000 219,500 5,096,029,300 Southern 0 391,900 0 House House Northern 16,100 242,000 4,087,193,300 Total Melbourne 6,800 1,220,100 9,183,222,600 Inner 0 59,900 0 Eastern 0 81,500 0 Western 16,000 37,500 585,892,500

Unit Southern 0 88,200 0 Northern 16,100 43,200 405,141,700 Total Melbourne 5,600 310,300 991,034,200 Source: SGS Economics & Planning

Based on average rate receipts for local government from land taxes this increased value of properties could add around $100 million dollars to the coffers of local governments every year. It would also add to higher stamp duties for the State government.

Confidential Copy 86

FIGURE 52. WESTERN RING ROAD IMPACTS TO HOUSE PRICES, 2011

Source: SGS Economics & Planning

Confidential Copy 87

FIGURE 53. WESTERN RING ROAD IMPACTS TO UNIT PRICES, 2011

Source: SGS Economics & Planning

Continuing Impacts The Western Ring Road provided increased accessibility from the Port of Melbourne to the west of Melbourne. This allowed many industrial freight and logistics firms to relocate to the west where there was cheap and available land. It is expected that the freeway will continue to provide these types of benefits well into the future. There is capacity for more firms to locate in the west, but still have access to the employment and services of the central core of Melbourne.

Confidential Copy 88

5.3 The City Loop

Impact to Employment As a result of numerous factors including data availability, a different methodology is required for the City Loop. The methodology has been developed to assess the impact of the City Loop using three techniques to measure the impact on employment. These three techniques are:

1. Density: By observing differences in employment densities within the CBD Grid an estimate of the impact the City Loop has had on the development can be made. That is, comparing sites surrounding the three City Loop stations with sites with poor access to stations. 2. Transport: By observing the levels of rail patronage over time an estimate of the capacity of the rail system without the City Loop can be estimated. Every additional job beyond that capacity figure within the CBD Grid can be attributed to the presence of the City Loop. 3. Share Analysis: In the 30 years prior to the opening of the City Loop the CBD Grid had been steadily losing employment share to the rest of Melbourne. Following the construction of the Loop, this downwards trend abated and in following years the CBD Grid began to increase its share of employment. The gap between the lowest share of employment and the CBD Grid’s share has been used to estimate the impact.

It is not possible to attribute all the employment growth in the CBD Grid to the City Loop as range of other factors (such as initiatives set out in the City of Melbourne’s 1985 Melbourne Strategy Plan) which have had an impact on the level of employment within the CBD Grid. Although, these methods should be considered as conservative as they assume that development only within the CBD Grid and not in other nearby parts of the city (East Melbourne, City North) can be attributed to the City Loop.

As with previous analyses in this section, the amount of additional employment new to Melbourne has been varied from 100%, 50% and 0%. This assumes varying levels of new employment in Melbourne, with 0% additional employment assuming jobs have been redistributed throughout Melbourne with total employment fixed at base case levels.

The table below presents the results of the three techniques in 2011.

TABLE 27. IMPACT OF CITY LOOP 2011, EMPLOYMENT SCENARIOS Jobs Density Share Analysis Transport capacity

100% Additional Employment 74,000 54,500 29,500 50% Additional Employment 37,000 27,250 14,750 0% Additional Employment - - - Source: SGS Economics & Planning

Example of Business Response to InfrastructureInfrastructure InIn ththee llateate 11990s990s anan intinternationalernational finfinancialancial seservicesrvices firm MaMazzeizzei & BBurtonurton basedbased in CanaryCanary Wharf in London was considerinconsideringg ooptionsptions for oopeningpening an Asia Pacific office. The companycompany shortlisted SySydneydney and MeMelbournelbourne as possiblepossible locations.locations. TheThe keykey criteria was centeredcentered aroundaround strongstrong bbusinessusiness opportunities,opportunities, affordable office spacespace and ggoodood access to a larlargege labour force via ppublicublic transtransport.port. A site ververyy simisimilarlar to ttheirheir London bbase.ase. WhWhileile SSydneyydney was hhighlyighly rated due to its llargearge financial services clustercluster,, the hihighgh ppricerice of office spacespace and ppooroor transtransportport links into the CBD concerned the company.company. Melbourne’s growinggrowing superannuationsuperannuation industryindustry was seen as a growthgrowth sector for MazzeiMazzei & BurtonBurton. Collins Street, with its pproximityroximity to Parliament Station was a far more attractiveattractive location than the SydneySydney CBD could offer. The comcompanypany relocated a dozen of their staff to Melbourne and employedemployed another dozen local staffstaff..

Confidential Copy 89

It is also possible that the urban system of Melbourne may have been able to adapt and accommodate these jobs elsewhere in Melbourne. That is, more suburb development or the development within the CBD Grid near Southern Cross or Flinders Street would have been of a larger scale. Given the contrasting nature of CBD Grid jobs and suburban jobs and the nature of the development industry it is unlikely that this would have been the case. However, to highlight the potential impact two additional scenarios have been constructed.

Table 28 presents the benefits for each scenario for the City Loop. They range from $1.27 billion (assuming that no jobs were lost from Melbourne but merely a redistribution of jobs across Melbourne) to $10.42 billion (assuming that the City Loop facilitated 74,000 jobs which would not have located in Melbourne otherwise).

TABLE 28. BENEFITS OF CITY LOOP, 2011 ($BILLION) Jobs Density Share Analysis Transport capacity

100% Additional Employment 10.42 7.68 4.88 50% Additional Employment 6.31 3.83 3.08 0% Additional Employment 3.18 1.92 1.27 Source: SGS Economics & Planning

These benefits are flowing from three sources, the first is agglomeration economies increasing labour productivity, the second is the impact of having more jobs in Melbourne and the third is a move to less productive employment. The third benefit only applies in scenarios where jobs are being reallocated across Melbourne.

Of the $10.42 billion benefit, $1.21 billion comes from improvements in the relative accessibility of particular locations which have been impacted by the project. This is created by improvements in labour productivity in these more accessible areas causing an increase to gross value added (GVA) across metropolitan Melbourne. The remaining $9.21 billion comes from the additional GVA generated by having 74,000 extra jobs within Melbourne.

Of the $6.31 billion benefit, $2.01 billion comes from improvements in the relative accessibility of particular locations which have been impacted by the City Loop. This is created by improvements in labour productivity in these more accessible areas causing an increase to gross value added (GVA) across metropolitan Melbourne. $3.04 billion comes from the additional GVA generated by having 37,000 extra jobs within Melbourne. The remaining $1.26 billion is from having an increased number of jobs in the CBD Grid where there are more highly productive industries.

Interestingly, the labour productivity benefit for the 100% additional employment scenario is around the same size as the 0% additional employment scenario. This can be interpreted as the impact of scattering jobs across Melbourne rather than in the CBD Grid has roughly the same impact as completely removing the jobs from the Melbourne. That is, dispersing jobs across Melbourne away from the CBD has a significant impact on all jobs in Melbourne.

The $1.21 billion increase in GVA would be split very roughly 50-50 between high profits and higher wages. Which working off an average tax rate of 32% would result in additional tax receipts for the Commonwealth of $384 million.

TABLE 29. CITY LOOP DETAILED BENEFITS FOR DENSITY SCENARIO, 2011 ($BILLION) Jobs Total Labour Productivity More jobs M2MPJ

100% Additional Employment $10.42 $1.21 $9.21 $- 50% Additional Employment $6.31 $2.01 $3.04 $1.26 0% Additional Employment $3.18 $1.27 $- $1.91 Source: SGS Economics & Planning

Confidential Copy 90

Increase in Land Value The method used to estimate the land value impacts of the City Loop project is the same as used for the CityLink and Western Ring Road projects, and has been outlined in Section 4.8. The effective job density change used in this analysis is under the Density Scenario. The average uplift in median house prices across Melbourne was estimated to be $11,900 as a result of the project. Using estimates from the Housing Development Data (HDD) survey conducted in 2009, the total number of houses across Melbourne was 1.22 million. This translated to a once off total impact on the economy of $13.2 billion.

These estimates are presented in Table 30 along with the results for units. The total uplift in house prices in the eastern region was estimated to be $4.21 billion calculated as the number of dwellings in this region (363,200) multiplied by the average median house price uplift. This was done at the suburb level as each suburb in the eastern region had median house prices varying around the average of $11,300. As shown, the inner, eastern and northern regions experience the largest impact to house and unit prices.

TABLE 30. CITY LOOP LAND VALUE IMPACTS Housing Region Average Uplift to Median Number of Dwellings Total Uplift Type House Prices

Inner 41,800 3,500 182,469,700 Eastern 11,300 363,200 4,214,310,000 Western 14,400 219,500 2,821,606,700 Southern 8,300 391,900 2,999,094,400 House House Northern 12,800 242,000 3,000,117,100 Total Melbourne 11,900 1,220,100 13,217,597,900 Inner 36,700 59,900 2,337,505,700 Eastern 9,800 81,500 1,671,253,800 Western 19,600 37,500 1,008,795,300

Unit Southern 6,900 88,200 1,397,254,100 Northern 31,700 43,200 1,506,834,900 Total Melbourne 15,100 310,300 7,921,643,800 Source: SGS Economics & Planning

Walk Time Impacts A broader analysis was undertaken to understand the ‘on the ground’ impacts of the City Loop. Walk times were calculated from key City Loop stations along with Southern Cross, Flinders Street and North Melbourne and Jolimont. The analysis shows the dramatic escalation in walking time in the absence of the City Loop, and provides further evidence on how accessibility was improved through the project, shaping the CBD. See Figure 54 for results which show reduced accessibility in the north of the CBD.

Example of Business Response to InfrastructureInfrastructure DaviesDavies & RawnsRawnsleyley were engineersengineers settingsetting upup a new consultingconsulting firm in the earlyearly 1990s. TheyThey plannedplanned to pproviderovide services to llocalocal bbuilders.uilders. TTheyhey were considerinconsideringg a susuburbanburban llocationocation in Melbourne, but the full openingopening of the CityCity LoopLoop in 1985 had spurredspurred a rangerange of office developmentsdevelopments in the northern partpart of the CBD. TheyThey knew the area well and saw that it was an areaarea oonn tthehe uup.p. DesDespitepite the rent beingbeing a bit more expensiveexpensive thetheyy rented a small office on Latrobe Street. WorkingWorking in the CBD, thetheyy qquicklyuickly came into contact with a rangerange of miningmining services firms whowho hhadad cclientslients in llargearge mininminingg comcompanies.panies. DaviesDavies & RawnsRawnsleyley qquicklyuickly rearealizedlized tthathat tthehe mininminingg sector pprovidedrovided far more income than local builders could pproviderovide them. DurinDuringg their first decade theythey steadilysteadily bbuiltuilt a stronstrongg cclientlient bbasease in AustraAustralialia pprovidingroviding services across tthehe countrcountry.y. TTheyhey were well pplacedlaced to take advantageadvantage of the miningmining bboomoom durinduringg tthehe 2000s and expandedexpanded operationsoperations internationally.internationally. This ggeneratedenerated stronstrongg exexportport earninearningsgs for Victoria and AustraliaAustralia..

Confidential Copy 91

FIGURE 54. WALKING ACCESSIBILITY – CITY LOOP VS NO CITY LOOP

Confidential Copy 92

5.4 Box Hill Central Activities Area development

Impacts to Employment and Housing The impacts to employment were estimated based on the case study comparison of Box Hill and Preston. A new level of employment by industry was estimated for Box Hill to simulate a ‘do nothing’ scenario. This lower level of employment was determined as the difference between employment in Box Hill and Preston. Figure 55 presents total employment from 1996 to 2011 in both centres, illustrating the converging levels of employment from 1996 onwards.

Another difference between the two centres is the share of total SLA employment each centre comprised. Figure 56 shows the share of employment in the Box Hill centre of total Whitehorse – Box Hill employment from 1996 and 2011. Over this time period it increased its share of employment as it became a more central location for the surrounding region. In contrast Preston centre’s share of the Darebin – Preston SLA has fallen over the same time period as other centres in surrounding areas become more dominant.

It can be reasoned that if the government intervention had not taken place in the early 1990’s Box Hill would not have experienced growing levels of employment. It would instead have an employment outcome similar to that of Preston. Therefore the impact to employment in 2011 is estimated to be approximately 7,700 jobs. This concentration of employment in the activity centre across a range of industries acts as a proxy for the access to services available to residents living in the surrounding region. As an example, additional health care employment in Box Hill represents a positive outcome for the local community who now have access to a greater number of doctors and health care services.

FIGURE 55. EMPLOYMENT IN BOX HILL AND PRESTON, 1996 TO 2011

Source: SGS Economics & Planning

Confidential Copy 93

FIGURE 56. CENTRES SHARE OF SLA EMPLOYMENT

Source: SGS Economics & Planning

The impacts to housing and population in Box Hill as a result of the project are harder to delineate for a number of reasons. These include data availability, historic household characteristics of Box Hill activity centre, and the competition for land between employment and residential uses. A similar analysis as completed for households was conducted for dwellings, comparing the time series of households over the last 15 years in Box Hill and Preston. Figure 57 presents the average annual growth rate for each five year period from 1996 to 2011 for households in Box Hill and Preston activity centres. Unlike employment, in 1996 Preston contained a greater number of households than Box Hill reflecting the greater lever of employment land in Box Hill. However household growth in Box Hill now eclipses that of Preston, with the gap in growth widening in more recent years. This is illustrative of in-centre housing developments, including apartments targeted towards students of Box Hill TAFE.

Whilst a variety of factors can be attributed to this stronger household growth, it can be reasoned that if the government interventions of this project did not occur household growth in Box Hill would be lower. This new growth trajectory in Box Hill would see it have more households than Preston, even though the land area of the centre is much smaller. This difference in land area is due to the shape of each activity centre, with Preston being more linear, based along High Street.

The competition for land uses should also be noted as having an impact upon the differing levels of employment and households arising as a result of this project. Compared to Preston, Box Hill has a greater amount of employment land allowing for more development to occur. This comes at the expense of additional households which instead choose to locate outside of the activity centre.

Confidential Copy 94

FIGURE 57. AVERAGE ANNUAL GROWTH RATE OF HOUSEHOLDS, PRESTON & BOX HILL

Source: SGS Economics & Planning

Economic Expansion The agglomeration model has been used to estimate the economic impact of investments into Box Hill, with the change in EJD estimated using the base travel time matrices and share of public transport use, with altered employment levels in some SLAs. This altered employment was estimated by applying the difference in level of employment between Box Hill and Preston to the Box Hill SLA. This gives a proxy for the potential outcome in Box Hill if government intervention had not occurred. To maintain a fixed Melbourne total of employment it was assumed that the extra jobs removed from Box Hill would be evenly distributed to the surrounding Eastern Melbourne SLAs. Figure 58 illustrates these estimated changes by SLA across Melbourne in 2011.

Confidential Copy 95

FIGURE 58. BOX HILL IMPACTS TO EFFECTIVE JOB DENSITY, 2011

Source: SGS Economics & Planning

Since this project is largely a land use project that did not have a significant impact on the accessibility of Box Hill the economic benefit comes solely from the increased number of jobs. When the labour productivity and EJD elasticities are applied to the project induced changes in EJD, the total impact on metropolitan GVA was estimated to be $500.8 million in 2006 and $586.7 million in 2011. Most of this benefit arises from benefits specific to the SLA containing the Box Hill centre and surrounds.

Confidential Copy 96

Table 31 presents the benefits by industry in 2006 and 2011 for all of metropolitan Melbourne. The largest benefit is forecast to flow from Health Care & Social Assistance ($241.6 million), Financial & Insurance services ($111.3 million) and Public Administration & Defence ($99.0 million). Figure 59 illustrates the geographical impact of the Box Hill project to SLA gross value added in 2011.

TABLE 31. BOX HILL IMPACTS ON INDUSTRY GROSS VALUE ADDED ($), MELBOURNE Industry 2006 Benefit 2006 Share of Benefit 2011 Benefit 2011 Share of Benefit

Agriculture, Forestry & Fishing -31,900 0.0% -30,200 0.0% Mining 200 0.0% -500 0.0% Manufacturing -16,980,900 -3.4% -13,908,500 -2.4% Electricity, Gas, Water & Waste 7,361,400 1.5% 8,613,700 1.5% Construction -17,799,300 -3.6% -21,339,500 -3.6% Wholesale Trade -24,765,400 -4.9% -26,517,700 -4.5% Retail Trade 10,045,300 2.0% 10,456,700 1.8% Accommodation & Food 6,568,100 1.3% 7,832,600 1.3% Transport, Postal & Warehousing -62,571,400 -12.5% -69,030,300 -11.8% Information Media & Telecom. 12,114,800 2.4% 7,786,000 1.3% Financial & Insurance 86,818,600 17.3% 111,298,700 19.0% Rental, Hiring & Real Estate 6,851,200 1.4% 8,044,100 1.4% Professional, Scientific & Technical 73,444,700 14.7% 86,738,400 14.8% Administrative & Support 96,433,000 19.3% 96,225,900 16.4% Public Admin & Safety 96,602,000 19.3% 99,045,800 16.9% Education & Training 19,296,200 3.9% 24,166,200 4.1% Health Care & Social Assist. 195,562,400 39.0% 241,662,900 41.2% Arts & Recreation 8,706,000 1.7% 12,061,000 2.1% Other Services 3,173,900 0.6% 3,601,900 0.6% Total 500,828,900 100% 586,707,200 100% Source: SGS Economics & Planning

Confidential Copy 97

FIGURE 59. BOX HILL ADDITIONAL EMPLOYMENT BENEFITS, 2011

Source: SGS Economics & Planning

Confidential Copy 98

6 KEY CONCLUSIONS

6.1 Summary of findings

It should be noted that conceptualising and quantifying economic benefits of these types for transport projects is still a new field. As more projects are appraised and the methodology is refined and debated within the transport appraisal and cost benefit communities there is likely to be on-going refinements to the methodology.

Of the four projects the City Loop has the largest estimated benefit ($10.4 billion), followed by CityLink ($9.0 billion), the Western Ring Road ($2.6 billion) and Box Hill ($0.6 billion). The transport projects have had a far greater impact on larger areas of the city than land use projects in Box Hill. The period of time the project has been in operation is also an important factor, with City Loop being constructed (and planned) earlier than the other projects. Figure 60 shows the stream of benefits for the three transport projects since their opening.

CityLink had a rapid ramp up in its benefit stream. The project improved the accessibility of large sections of Melbourne. A large number of small scale economic development events were able to occur fairly quickly across a large area with a diverse economy. The last three or four years have been impacted by slower overall economic growth so it is difficult to fully know if the growth in benefits from CityLink has reached a plateau or is merely experiencing a natural pause resulting from broader economic trends.

Over the life of the City Loop there are several periods when broader economic trends have impacted on the stream of benefits. As shown in Figure 60 Years 11-15 corresponded of the recession of the early 1990s and years 20-23 correspond with the economic slowdown following the introduction of the new tax system in 1999-2000. Those two plateaus also correspond with significant development occurring in Southbank (in the early 1990s) and Docklands (early 2000s). However, following these pauses the presence of the City Loop allowed on-going investment and facilitated a range of economic benefits. The most recent years have been impacted by slowing economic growth.

Confidential Copy 99

FIGURE 60. BENEFIT STREAM ACROSS TIME $BILLION (EXCLUDING BOX HILL)

Source: SGS Economics & Planning

The Western Ring Road exhibits a similar pattern to CityLink but the ramp up period was not as pronounced and the levelling off has been more significant over the past few years. This is driven by the different composition of benefits flowing from the two projects. CityLink has enabled a labour productivity improvement of $1.4 billion and freight improvements of $81 million. The Western Ring Road has a much more equal split of benefits flowing from these two sources. The respective benefits are labour productivity of $228 million and freight improvements $149 million. Figure 61 presents a breakdown of these benefits to Victorian gross value added for each of the four projects.

Confidential Copy 100

FIGURE 61. SUMMARY OF PROJECT BENEFITS IN 2011, ($MILLION)

Source: SGS Economics & Planning

This reflects the position of the two road projects within Melbourne. CityLink as the names suggests helped to link together the north and south east of Melbourne with the central core of the City. While the Western Ring Road serves the industrial zones of the north west of Melbourne.

TABLE 32. SUMMARY OF BENEFITS TO METROPOLITAN MELBOURNE, 2011 ($MILLION) Benefit CityLink Western Ring Road Box Hill City Loop

Project Conceptualisation 1969 1954 1954 1929 Project Completed 2000 1999 1992 1985 Productivity Improvements $1,395 $228 $0 $1,205 Move to More Productive Jobs $7,547 $2,216 $587 $9,214 Total GVA Uplift $9,023 $2,593 $587 $10,419 New Jobs 70,300 24,900 0 74,000 New Households 58,200 17,400 n.a. n.a. Freight Improvements $81 $149 n.a. n.a. Freight Travel Time Savings -0.8% -1.6% n.a. n.a. Human Capital $14 $5 $0 n.a. Land Value Improvements $29,646 $10,174 $0 $21,139 Source: SGS Economics & Planning

6.2 Areas for Further Refinement

The project has revealed that there is a need for improved data collection at the small area level. For example: - Detailed and easily accessible data on industrial and commercial land values would be helpful in assessment of future projects. - Data on productivity at the firm level. - Longitudinal data on individual’s lifetime labour earnings. - Improved historical travel time matrices.

Confidential Copy 101

- Interaction between transport and land use changes, for example how different types (public transport and road) of transport improvements are linked to increased density. - Transport project appraisals needs to evolve further, taking on more of the character of dynamic general equilibrium modelling. That is, they leave the full benefit of city shaping projects unaccounted for as they focus only on first round effects (as is the convention in CBAs). - Major strategic transport projects need to be conceptualised in the context of a preferred urban structure (e.g. creating the sort of city we want) rather than the traditional predict and provide thinking.

Discussions of the techniques and methods within the transport appraisal and cost benefit communities to gain a common understanding. Once a common understanding is reached, the method can be codified and applied to future projects.

6.3 Conclusions

These projects have fulfilled the important roles they were designed to achieve, but benefits go far beyond those solely associated with transport improvement.

These projects have helped to reshape the economic geography of Melbourne. The City Loop has transformed the northern section of the CBD Grid, while the north west of Melbourne has been opened up for both industrial and residential development by the Western Ring Road. CityLink has provided heightened connectivity between the central core and north and south eastern parts of Melbourne. The data, statistical analysis and anecdotal evidence all support this and can been seen by anyone familiar with the history of Melbourne.

This report has also highlighted something which may not be as obvious. These projects allowed the overall economy of Melbourne and Victoria to expand. In the early 1990s before these projects commenced or begun to fully be appreciated, the ABS population projections were anticipating a Victoria with 5.0 million people in 2011. In 2011 over 5.6 million people lived in Victoria9.

The economic development these projects generated led employment, population and income growth to be significantly higher than the trends that had been anticipated. This higher growth provided all levels of government with hundreds of millions of dollars more in taxation revenue than would have been the case without this transport infrastructure. It is of significant note that government has enacted mechanisms designed to capture some of the uplift in land value which would flow from the project to assist with funding.

In moving forward in infrastructure planning, it is important to acknowledge the potential city shaping effects of transport projects, which can have substantial impacts, over and above the immediate relief they can bring to rising congestion and travel time.

9 http://www.ausstats.abs.gov.au/ausstats/free.nsf/0/B38E74C292E50B68CA257225000494C5/$File/32220_1993.pdf

Confidential Copy 102

APPENDIX 1: DATABASE DEVELOPMENT

To assess the economic impacts of infrastructure projects in Melbourne, a database was developed that contained the following variables across the metropolitan area from 1986 to 2011 at the smallest feasible geographic level (Statistical Local Area):

 Travel time by origin and destination zones (by mode)  Population and dwellings  Employment by major industry group (e.g. manufacturing, retail, professional services, etc.)  Labour force participation  Labour productivity (value added per hour by industry)  Effective job density  Property values,  Relative accessibility of specific industries (to priority locations given their value chain).

Each of these variables, how they were developed and the potential insights they provide into the long term impacts of projects are described below. The way in which these variables were prepared is detailed in the Technical Appendix of this report. The way the variables have been applied to assess the impacts of the City Loop, Western Ring Road, CityLink and investments into Box Hill Activities Area has been described alongside results in the Impact Assessment Section (Chapter 5).

Travel Time Matrices

A travel time matrix shows the time it would take to travel from one Statistical Local Area (SLA) to another by car and public transport, for peak (morning and afternoon) and off peak times. Travel time matrices for 1996, 2001, 2006 and 2011 were sourced from the Department of Transport’s Melbourne integrated transport model (MITM).

A separate transport modelling program was investigated for use in assessing road transport projects (CityLink and Western Ring Road) called Route Finder. This program uses network analysis to plot the shortest available route given a set road network. Any particular road can be turned off and the simulation run to determine the alternate travel times between SLAs. This model was run at the higher order road level, with the CityLink roads (and Western Ring Road separately) turned off and then compared to the simulation with all roads turned on. The percentage difference in travel times with and without the project was then applied to the base travel time matrix from 1996 to estimate a project case travel time matrix.

Confidential Copy 103

Population and Dwellings

Population and dwelling estimates for all SLAs in Melbourne from 1981 to 2011 have been estimated using ABS data. From 1991 to 2011 the Estimated Resident Population (ERP) dataset released by the ABS has been used. Estimates from 1981 to 1991 have been sourced from the Suburbs in Time database produced by DPCD which compiles the ABS Census data at a suburb level. The total population in metropolitan Melbourne and Victoria back to 1981 has been taken from the Australian Historical Population Statistics (cat. no. 3105.0.65.001).

This dataset contains estimates of employment in major capital cities and centres and for the States and Territories from 1901 onwards. Figure 62 shows the outputs of this dataset for Melbourne and Victoria.

FIGURE 62. POPULATION IN MELBOURNE AND VICTORIA, 1981 TO 2011

Source: ABS Estimated Resident Population, cat. no. 3105.0.65.001 and cat. no. 3218.0

The annual average growth rates in population for a selected number of SLAs is shown Table 33. Of particular interest, is the significant decline in population that occurred from 1986 to 1991 in the SLAs of Melbourne – Inner, Southbank/Docklands and Remainder, along with decline in Brimbank – Sunshine, Banyule – North, Whitehorse – Box Hill and Glen Eira – Caulfield. From 1991 onwards however, the SLAs of Melbourne – Inner and Southbank/Docklands experienced much higher levels of growth than other SLAs.

Confidential Copy 104

TABLE 33. SELECTED SLAS AVERAGE ANNUAL GROWTH RATES OF POPULATION SLA 1986-1991 1991-1996 1996-2001 2001-2006 2006-2011

Melbourne - Inner -30.1% 6.5% 29.6% 13.9% 4.8% Melbourne - S'bank-D'lands -28.1% 73.9% 15.3% 25.7% 6.2% Melbourne - Remainder -11.9% 1.5% 4.6% 4.0% 3.3% Port Phillip - St Kilda 3.3% 1.3% -0.1% 1.7% 1.4% Port Phillip - West -0.7% 1.4% 3.3% 3.3% 2.1% Stonnington - Prahran 1.3% 0.7% 0.4% 1.3% 1.4% Yarra - North 1.0% 1.3% 0.4% 1.3% 1.8% Brimbank - Sunshine -4.6% 0.0% 0.0% 1.4% 3.1% Wyndham - North 1.7% 3.8% 2.6% 3.7% 6.1% Darebin - Preston -0.5% -0.4% 0.1% 1.0% 1.2% Hume - Broadmeadows 0.9% 0.8% -0.1% -0.8% 0.5% Whittlesea - South-West 0.2% 1.7% 0.7% -0.6% 0.8% Boroondara - Camberwell N. 4.5% 0.2% 0.7% 0.6% 0.9% Monash - South-West -0.2% 0.2% 0.7% 1.5% 1.4% Whitehorse - Box Hill -3.6% 0.3% 0.8% 1.0% 1.1% Gr. Dandenong - Dandenong 5.1% -0.4% -0.2% 0.1% 1.2% Source: ABS Estimated Resident Population, SGS Economics & Planning

Employment by industry

Estimates of employment by industry from 1981 to 2011 have been compiled using a variety of data sources and informed assumptions to create a consistent time series.

Employment by industry estimates for the years from 1996 to 2011 have been produced using shares from the ABS Census years (1996, 2001 and 2006) and interpolated for the years in-between. The ABS Census Journey to Work data has been used to estimate employment in each SLA for 1996, 2001, and 2006. However, due to undercounting in this dataset, the estimates for Melbourne were benchmarked to annual average employment estimates for each industry from the Labour Force Survey for each year. An adjustment has been made to the Labour Force Survey to account for people who live in Regional Victoria but travel to Melbourne for work. City of Melbourne Census of Land Use and Employment (CLUE) data was used to adjust the Census Journey to Work industry shares for the most recent years (available for 2002, 2004, 2006, 2008 and 2010).

Employment estimates for the years from 1986 to 1996 were generated using a variety of data sources and assumptions. Numbers for the CBD and City of Melbourne (LGA) were sourced from the City of Melbourne Strategy Plan 1985. The Suburbs in Time database was used to obtain estimates of SLA employment in 1986. SLA shares of employment were then estimated for each year and applied to a total Melbourne figure. This estimate of metropolitan Melbourne employment back to 1985 came from the ABS Labour Force Survey. The estimates from 1981 to 1985 were estimated using the average annual growth rates from historical Censuses. Figure 63 presents estimates of employment in the CBD and City of Melbourne from 1981 to 2011. As shown, employment is fairly flat until 1991/2, where a drop off occurs reflecting the recession. From 1991/2 to 2011 employment more than doubles in the Melbourne LGA, reaching almost 500,000 jobs. The approximate share of the City of Melbourne’s jobs located in the CBD also declines over this time, owing to development in Southbank, Docklands, St Kilda Road, and to a lesser extent, Carlton.

Confidential Copy 105

FIGURE 63. HISTORICAL EMPLOYMENT IN THE CBD AND CITY OF MELBOURNE

Source: ABS, SGS Economics & Planning

Figure 64 presents estimates of industry shares of total employment in metropolitan Melbourne in 1996 and 2011, showing the decline in Manufacturing and Wholesale trade employment. Also prominent is the rise in knowledge intensive services such as Professional, Scientific & Technical Services over the past 15 years. Table 34 presents the average annual growth rates in employment for selected SLAs for each five year period from 1986 to 2011. Of note is the strong growth in Port Phillip – West, Yarra – North, Hume – Broadmeadows and Knox – South. The low starting base of these SLAs has contributed to the large growth rates demonstrated. In the past five years growth has been strongest in the three SLAs of Boroondara – Camberwell N and Glen Eira – Caulfield.

Confidential Copy 106

FIGURE 64. METROPOLITAN MELBOURNE EMPLOYMENT BY INDUSTRY, 1996 AND 2011

Source: ABS, SGS Economics & Planning

TABLE 34. SELECTED SLAS AVERAGE ANNUAL GROWTH RATES OF TOTAL EMPLOYMENT SLA 1986-1991 1991-1996 1996-2001 2001-2006 2006-2011

Melbourne - Inner -1.5% 2.0% 3.8% 2.7% 3.6% Melbourne - S'bank-D'lands 3.9% 11.0% 3.3% 7.8% 8.8% Melbourne - Remainder 3.9% 11.0% 0.8% 0.3% 5.8% Port Phillip - St Kilda 2.5% -0.4% 3.2% 1.1% 2.6% Port Phillip - West 23.9% 8.7% 2.9% 1.3% 2.2% Stonnington - Prahran 5.4% 1.9% 1.2% -0.2% 2.1% Yarra - North 13.3% 5.8% -0.2% 0.3% 1.5% Brimbank - Sunshine -2.7% -6.3% 3.6% 3.6% 1.9% Wyndham - North 3.4% 0.2% 6.2% 6.2% 1.8% Darebin - Preston 3.9% 0.8% -3.0% 1.0% 1.8% Hume - Broadmeadows 10.8% 4.6% 1.5% 1.6% 0.8% Whittlesea - South-West -0.2% -3.0% 2.8% 3.9% 1.5% Boroondara - Camberwell N. -1.3% -4.4% 0.9% 2.2% 2.9% Monash - South-West 4.8% 1.9% 0.3% 0.5% 1.9% Whitehorse - Box Hill 1.7% -1.0% 3.1% 1.7% 2.2% Gr. Dandenong - Dandenong 7.1% 3.1% -0.1% 1.5% 1.4% Source: ABS, SGS Economics & Planning

Labour force participation

Over the past 30 years the labour force participation rate in Melbourne has steadily increased from just under 63% in 1981 up to 67% in 2011. Given the various macroeconomic conditions such as the recession of the early 1990’s and more recent global financial crisis the participation rate has varied significantly during this time. Figure 65 presents the annual participation rate for Melbourne from 1981

Confidential Copy 107

to 2011 illustrating this variability in labour force participation. This data has been sourced from the ABS Labour Force Survey.

FIGURE 65. ANNUAL PARTICIPATION RATE, MELBOURNE

Source: ABS Labour Force Survey Cat. No. 6202.0

Labour productivity

Labour productivity is calculated by dividing the Gross Value Added (GVA) for an industry by the total number of hours worked in that industry.

ܩܸܣ௜ ܮܲ௜ ൌ ௜݀݁݇ݎ݋ܹݏݎ݋ݑܪ

Where: ܮܲ௜ is the Labour Productivity for zone i ܩܸܣ௜ is the Gross Value Added for zone i ௜is the number of hours worked for zone i݀݁݇ݎ݋ܹݏݎ݋ݑܪ

For more detail on how labour productivity is estimated, please see the Technical Appendix.

Labour productivity by industry estimates were first produced at a state and capital city level for Victoria and Melbourne. These were broken down into a small area level (SLA) and estimated for each year from 1996 to 2011. The employment by industry dataset was used to estimate gross value added at a small area level for the non-Census years.

Figure 66 shows the labour productivity for metropolitan Melbourne from 1996 to 2011. Table 35 presents the labour productivity at five yearly intervals from 1996 to 2011 for the inner, middle, outer and growth area regions, developed from an aggregation of SLAs.

Most notably, the labour productivity in the inner region has increased significantly in the past 15 years. This is compared to modest increases in the middle and outer regions. The growth region has experienced a large growth rate which is coming off a low base in 1996 as these areas have become

Confidential Copy 108

more developed. The inner region has a labour productivity of $74.3 per hour worked in 2011 which was the largest of the regions. It was also more productive than the Melbourne average overall ($65.2).

FIGURE 66. MELBOURNE LABOUR PRODUCTIVITY

Source: ABS State Accounts cat. no. 5220.0

TABLE 35. LABOUR PRODUCTIVITY BY REGION Region 1996 2001 2006 2011 Inner 51.8 61.6 68.3 74.3 Middle 41.7 44.9 47.5 47.8 Outer 33.4 38.5 44.2 44.1 Growth 29.2 35.9 45.7 45.0 Total Melbourne 49.9 56.8 62.9 65.2 Source: ABS State Accounts cat. no. 5220.0 and SGS Economics & Planning

Industry labour productivity has also varied over the past 15 years, with the level of change varying across industries. Figure 67 presents the labour productivity by industry for metropolitan Melbourne in 1996 and 2011. On the whole most industries experienced an increase in labour productivity. This was most pronounced in the Financial & Insurance industry which grew from $84.1 per hour worked in 1996 to $155.9 per hour worked in 2011.

Confidential Copy 109

FIGURE 67. METROPOLITAN MELBOURNE INDUSTRY LABOUR PRODUCTIVITY, 1996 AND 2011

Source: ABS State Accounts cat. no. 5220.0 and SGS Economics & Planning

Effective Job Density

To measure agglomeration within an area, the variable of Effective Job Density (EJD) is used. EJD is the sum of employment in an area and the employment in all other areas divided by the travel time in reaching these external jobs, with travel time weighted for transport mode. Essentially it demonstrates the density of employment in a location, accounting for accessibility.

To calculate EJD, the level of employment relative to the time taken to gain access to that employment and the mode split that is currently experienced by those workers in their travel to employment is used. The formula used to calculate EJD at an SLA level is presented below.

ܲܶܯ݋݄݀݁ܵܽݎ݁௝ ൈܧ݉݌௝ ሺͳ െ ܲܶܯ݋݄݀݁ܵܽݎ݁௝ሻൈܧ݉݌௝ ܧܬܦ௜ ൌ෍ቆ ൅ ቇ ݒ݈݁ܶ݅݉݁௜௝ܽݎܸܶܲ ݒ݈݁ܶ݅݉݁௜௝ܽݎܶܶܲ ௝

Where:

ܧܬܦ௜= Effective Job Density for SLA i ܲܶܯ݋݄݀݁ܵܽݎ݁௝ = per cent of work trips which involve public transport for SLA j ܧ݉݌௝= number of jobs/employment within SLA j ݒ݈݁ܶ݅݉݁௜௝ = time it takes to travel on public transport from SLA i to SLA jܽݎܶܶܲ ݒ݈݁ܶ݅݉݁௜௝ = time it takes to travel by private vehicle from SLA i to SLA jܽݎܸܶܲ

The public transport mode share and public transport and private vehicle travel times come from Census data and the MITM respectively.

Estimates of EJD have been calculated for 2011, as shown in Figure 68. This series has been produced using 2011 travel time matrices and employment. The same method was used to estimate EJD for 1996, 2001 and 2006 using travel time matrices and employment for each of the corresponding years. A time series from 1981 to 2011 of EJD was then estimated using varying employment for the intervening years.

Confidential Copy 110

FIGURE 68. EJD IN MELBOURNE, 2011

Source: SGS Economics & Planning

Property values

The Melbourne property market has experienced a substantial increase in property sale prices over the 20 year period from 1991 to 2010. While prices remained relatively stable in the early 1990s, from the late 1990s the prices began to rise substantially each year and are continuing on an upwardly increasing trend for Houses, Units and Vacant Land.

House sale prices are increasing across the Melbourne area at an average rate of 7.7% per annum while units are increasing at a similarly high average rate of 6.9% per annum. As vacant land is becoming increasingly scarce within inner established areas of Melbourne, vacant land prices are also increasing at even higher average rate of 8.2% per annum.

The Inner ring of Melbourne experienced the highest prices for both Houses and Units, followed by the Middle and Outer ring10. The differences in price between the rings were relatively similar in the early

10 Inner Ring includes the municipalities of Melbourne, Yarra, Stonnington and Port Phillip. Middle Ring includes municipalities of Bayside, Banyule, Boroondara, Brimbank, Darebin, Glen Eira, Hobsons Bay, Kingston, Knox, Manningham, Maribyrnong,

Confidential Copy 111

1990s. However, the large increase in property prices has also resulted in a substantial increase in the differences in price between the three rings, with high demand for the Inner core.

In terms of the regions and rings, the Outer Western region of Melbourne has remained the cheapest place to buy a house or unit over the 20 year period. The Middle Southern region (for houses) and Middle Eastern region (for units) were the most expensive outside of the Inner area. The highest annual growth for Houses took place in the Inner region (9.1%), followed by the Middle Southern region (8.6%), while the lowest growth occurred in the Outer North (6.1%). For units the highest annual growth occurred in the Middle South (7.5%) and the lowest in the Outer West (5.6%).

Figure 69 presents the sale trend for houses by ring from 1992 to 2009. The Inner ring of Melbourne has remained the most expensive over this period, followed by the Middle and Outer rings. The early 1990s saw only a slight variation in price between the rings which has changed dramatically in more recent years. In 2009 there was a large difference between the three rings in house prices. This shows the increasingly level of preference and high demand for locations close to the CBD. In the Inner ring house property sales increased at an average of 9.1% p.a., while in the Middle ring they increased by 7.7% and in the Outer ring by 6.6%.

FIGURE 69. SALE TRENDS FOR HOUSES BY RING, 1992 TO 2009

Source: Valuer General, SGS Smoothed Series

The annual average growth rates (AAGR) presented above were determined based on input data from the Valuer General. The data was derived from actual sale records during the period 1991 to 2010, with some years having no sales recorded and thus no available data. To allow for this in the analysis of AAGR, suburbs that had no sales or less than five sales recorded over the 1991-2010 period were removed from the analysis. To determine the average annual growth rate, a moving average was calculated on the data. This smoothed out any short term price fluctuations, giving price data from 1992 to 2009.

Figure 70 illustrates the differences in median house prices depending on the distance of that suburb from the CBD. Each suburb has been plotted for 1990-91, 1999-00 and 2009-10. In the 15 years to 2009-

Maroondah, Monash, Moonee Valley, Moreland, and Whitehorse. Outer Ring includes Cardinia, Casey, Frankston, Hume, Melton, Mornington Peninsula, Nillumbik, Whittlesea, Wyndham and Yarra Ranges.

Confidential Copy 112

10, house prices grew much faster in absolute terms in the inner and middle suburbs located closer to the CBD. There is now a pronounced gradient of increasing prices with reducing distance to the CBD.

FIGURE 70. MEDIAN HOUSE PRICES AND DISTANCE FROM CBD

Source: SGS Economics & Planning

Land Value Impacts The projects examined in this report will also have had an impact on the land values of the areas affected by this major investment in infrastructure. To quantify the value of this benefit to the economy SGS has developed an approach using regression analysis. The statistical relationship between house prices (used as a proxy for land values) and effective job density (EJD) was estimated. It was found there was a positive relationship between these two variables. This indicates that in areas that are effectively more dense (higher EJD) in terms of employment or access to employment, house prices will be higher.

Regression analysis is a statistical technique used widely in econometrics to focus on the relationship between a dependent and independent variables. It is used extensively in prediction and forecasting models with continuing research being conducted in this area to develop more accurate forecasts. Historical data back to 1991 demonstrates an observed statistical relationship between house and unit prices and access to employment opportunities in Melbourne. It is therefore sensible to estimate the statistical relationship between these two variables. Hedonic price analysis is a form of regression analysis that can be used in relation to property values and the influence of various demand factors.

Due to data availability the relationship between commercial land values and EJD could not be estimated. However commercial land values will be affected to a greater extent by accessibility factors, including EJD, as commercial property developers require a significant amount of transport infrastructure to make a project feasible. These drivers of demand are similar for residential and commercial properties to a certain extent. The analysis conducted in this report focuses solely on the demand factor of accessibility and thus provides some guidance as to the impact of the projects on commercial property values in addition to residential values.

Whilst house prices are affected by a variety of factors, this method has controlled for the differences between house prices in the west and south east of Melbourne. A separate regression was run for house and unit prices. This type of analysis only examines the relationship between EJD and house prices and

Confidential Copy 113

does not attempt to model any of the macroeconomic factors that affect house prices. Table 36 presents the regression results for the coefficients used in this land value impacts analysis.

TABLE 36. LAND VALUE REGRESSION RESULTS Housing Type B0 (EJD) B1 (Constant) B2 (West Dummy) B3 (South East R squared Dummy) House 5.95 43,257 -36,158 146,924 0.42 Unit 4.02 -103,205 10,507 92,311 0.37 Source: SGS Economics & Planning

In order to estimate the uplift to house and unit prices the regression coefficients were applied to the percentage change estimated for each project. This determined a percentage uplift in house and unit prices arising from the change in EJD. This percentage change was then applied to the base estimates of median house and unit prices by suburb to determine a project case median house price. The total benefit to Melbourne was then estimated as the sum of the suburb median house price change multiplied by the number of dwellings in each suburb. Given the large variation in house prices the changes to median house prices were capped at the 95th percentile. For example, in the suburb of Armadale in Stonnington – Prahran where the median house price is $1.55 million, the uplift to EJD translates to an unrealistic increase in the median house price in this area. It is unlikely this is going to have this large of an effect on average across the entire suburb.

Comparison of Box Hill and Preston Activities Areas To understand the impact of government investment into the Box Hill Central Activities Area, it is worth conceptualising what the centre may look like and what its present-day role in the metropolitan economy may be like had it not been the subject of government intervention. To do this, this section provides a comparison with a similar centre which did not receive the same level of government investment.

In the 1954 Metropolitan Melbourne Planning Scheme, five district centres were defined: Box Hill, Preston, Footscray, Moorabbin and Dandenong. Among these centres, Box Hill and Preston shared the most similarities. However, these centres have followed different development trajectories and serve different roles in the metropolitan Melbourne today.

In the early 1950s, High Street, Preston, had 625 shops and a population of around 55,650 in its catchment area. Station Street in Box Hill had 632 shops and a population catchment of 52,000 to draw from (MMBW, 1954: p. 95). Both centres had their own railway station providing access to the CBD and beyond, and both were initially settled and subdivided at a similar time period.

The two centres also had similar levels of other infrastructure. The Box Hill Hospital opened in 1956, whilst the Preston and Northcote Community Hospital (PANCH) opened in 1960 (see images below). In the 1990s however, PANCH hospital closed and was replaced by the Northern Hospital in Epping, to serve the rapidly growing residential areas on the northern fringe.

Confidential Copy 114

FIGURE 71. BOX HILL HOSPITAL 1960S AND PANCH HOSPITAL 1950S

Source: http://dhe.darebin-libraries.vic.gov.au/encyclopedia.asp?id=971, http://www.slv.vic.gov.au/roseglass/gid/slv-pic-aab95816/1/rg008576

Both Activities Areas also have their own education facilities. The Box Hill Institute of TAFE was a product from the merger between the Box Hill Technical School for Girls and Women (est. 1924) and the Box Hill Technical School for boys (est. 1943). The Preston campus of the Northern Metropolitan Institute of TAFE was originally developed as the Preston College of TAFE in 1945. By 1951, the Preston College of TAFE was the biggest technical school in terms of enrolment in Victoria.

Today, Box Hill and Preston retain most of the same infrastructure and assets. Both are extremely well connected to the Principal Public Transport Network, both have education facilities and both have a sizable demographic catchment. However, the number of jobs supported by the two centres differs considerably. There are several historical reasons pointing to why this is the case. These include the aforementioned closure of PANCH, and also, the opening of Northland Shopping Centre, and more broadly, urban development which occurred beyond both centres towards the city fringe.

In 1966, Northland Shopping Centre opened in Preston West, some 3.5 kilometres to the west of High Street in Preston. This centre directly competed with High Street, Preston, by capturing retail employment and spend. It has expanded significantly since its construction, and is now a major shopping destination for Melbourne’s North.

Confidential Copy 115

FIGURE 72. NORTHLAND SHOPPING CENTRE, MYER ENTRANCE 1967

Source: http://nla.gov.au/nla.pic-vn3425010, Sievers, Wolfgang.

A broader reason pointing to why there are differences between the centres include development which has occurred further out, towards Melbourne fringe. Beyond Box Hill, the eastern suburbs are cut short by the Yarra Ranges, limiting significant Greenfield development. As such, the regional context in which Box Hill operates has largely remained the same, and it has continued to be a key centre for the east. In contrast, significant development has occurred to the north of Preston and the northern growth areas continue to expand. This has seen the region’s population dramatically grow in size, but has also seen the population centre for the region change. Centres like Epping are more accessible for many residents.

Notwithstanding the historical differences between Box Hill and Preston, the government investment poured into Box Hill has undoubtedly positioned it as a leading regional centre in Melbourne today. The following section examines employment in the centres to help conceptualise how Box Hill may have otherwise operated.

Confidential Copy 116

APPENDIX 2: TECHNICAL NOTES

Labour productivity estimation Labour Productivity by Industry (State & Capital City Level)

Labour productivity is calculated by dividing the Gross Value Added (GVA) for an industry by the total number of hours worked in that industry.

ܩܸܣ௜ ܮܲ௜ ൌ ௜݀݁݇ݎ݋ܹݏݎ݋ݑܪ

Where: ܮܲ௜ is the Labour Productivity for zone i ܩܸܣ௜ is the Gross Value Added for zone i ௜is the number of hours worked for zone i݀݁݇ݎ݋ܹݏݎ݋ݑܪ

The estimate of GDP for each capital city is derived from industry data published in the Australian National Accounts: State Accounts (cat. no. 5220.0) publication.

Sectors of employment are defined in the ABS Australia & New Zealand Standard Industrial Classification (ANZSIC 2006). Table 37 outlines each industry, with some examples of the activities which fit into each category, but it does not provide an exhaustive list of what each industry contains.

Confidential Copy 117

TABLE 37. SECTOR CLASSIFICATIONS Industry Contains Farming, Forestry, Fishing, Aquaculture Agriculture, Forestry & Fishing

Coal Mining, Gas Extraction, Mineral Mining, Exploration Mining

Food Products, Textiles, Wood & Paper, Petroleum, Metal, Machinery & Manufacturing Furniture Manufacturing

Supply of Electricity, Gas & Water, Waste Collection & Treatment Electricity, Gas, Water & Waste Services

Building, Heavy & Civil Engineering Construction & Services Construction

Material, Machinery, Motor Vehicle & Grocery Products Wholesaling Wholesale Trade

Motor Vehicle, Fuel, Food & Other Retailing Retail Trade

Accommodation, Cafes, Restaurants Accommodation &Food Services

Road, Rail, Water & Air Transport, Postage & Support Services Transport, Postal & Warehousing

Publishing, Broadcasting, Internet Services Providers, Motion Picture Information Media & Telecommunications Recording & Libraries

Banking, Superannuation & Insurance, Auxiliary Finance Financial & Insurance Services

Rental & Hiring Services, Property Operators & Real Estate Agents Rental, Hiring & Real Estate Services

Scientific Research, Architects, Legal, Accounting, Management, Professional, Scientific & Technical Services Consulting, Veterinary & Computer Related Services

Employment Services, Travel Agencies, Pest Control, Packaging Services Administrative & Support Services

Government Administration, Defence, Public Order & Regulatory Public Administration & Safety Services

School, Tertiary & Adult Education Education & Training

Hospitals, Medical Services, Allied Health, Residential Care & Child Health Care & Social Assistance Services

Museum, Parks & Gardens Operations, Creative & Performing Arts, Arts & Recreation Services Sports & Physical Recreation, Gambling

Automotive & General Maintenance, Personal Care, Funeral & Religious Other Services Services, Interest Groups

Source: ABS ANZSIC 2006

There are three approaches to measuring GDP; the Production approach (the sum of the GVA for each of the industries and taxes less subsides on products); the Expenditure approach (measures final expenditure on goods and services) and the Income approach (sum of income generated by all factors of production).

At the Australian level, the Production, Expenditure and Income approaches are averaged by the ABS to produce GDP. However, at the State level due to a lack of data on interstate trade, the Expenditure and Income approaches are combined and averaged with the Production approach.

Confidential Copy 118

In developing the GDP11 for each capital city the Production approach is used, firstly because of this lack of data on intrastate trade. Another major reason is that the data available to calculate the Production approach is more robust (and hence requires fewer assumptions to be made) than that available for the Expenditure or Income approaches. For each industry, wherever possible, the same data sources which have been used to produce industry Gross Value Added at the State level are used to produce industry Gross Value Added at the city level. Some of these data sources include:

 Agricultural Commodities: Small Area Data, Australia (cat. no. 7125.0)  Business Indicators, Australia (cat. no. 5676.0)  Manufacturing Industry, Australia (cat. no. 8221.0)  Regional Population Growth, Australia (cat. no. 3218.0)  Household Expenditure Survey (cat. no. 6530.0)  Education and Training Experience (cat. no. 6278.0)  Labour Force, Australia, Detailed, Quarterly (cat. no. 6291.0.55.003)

In order to maintain consistency with the wider National Accounts, the Production approach estimate of Gross City Product is benchmarked to the Gross State Product using a statistical discrepancy method. Via the use of the implicit price deflation technique, the Chain Volume Measures of the Gross City Product are converted into Current Prices. This method overcomes the non-additivity issue with the Chain Volume Measure.

The estimates of hours worked are derived from Information Paper: Implementing New Estimates of Hours Worked into the Australian National Accounts, 2006 (5204.0.55.003) which provides the total hours worked within the economy for 2004-05. The index of total hours worked from the Australian System of National Accounts, 2007-08 (cat. no. 5204.0) can be used to advance the 2004-05 estimate for the years between 2005-06 and 2007-08. This Australian ‘total hours worked’ figure can then been allocated for each industry in each capital city based on its share of total hours worked from the Labour Force, Australia, Detailed, Quarterly (cat. no. 6291.0.55.003).

There are several advantages of using this measure of labour productivity. It is built on the National Accounts framework. This allows the agglomeration benefits to be viewed in the context of the wider economy. This includes the Australian, State and City economies. The National Accounts also provide a clear methodology for measuring economic activity and labour productivity. Direct comparisons of the benefits of agglomeration can be made between and within Australian cities. It should be noted that, in conducting this analysis, industries outside the Australian Bureau of Statistics Market Sector are included.

As an example, Table 38 shows how the larger cities of Melbourne and Sydney have higher labour productivity than the smaller cities. This provides the first clear indication that the size of the economy of a city can provide improved outcomes for labour productivity. The industry mix within each city would also have an influence on the outcome. That is, more productive industries may tend to locate in particular cities.

TABLE 38. MAJOR CAPITAL CITIES LABOUR PRODUCTIVITY ($ OF GROSS VALUE ADDED PER HOUR WORKED) Sydney Melbourne Brisbane Adelaide Perth

1999 67.3 56.1 56.5 55.7 61.9 2004 71.2 61.9 61.3 59.5 71.6 2009 73.7 66.4 62.2 63.2 64.0 2011 75.7 65.2 64.0 64.5 71.1 Source: SGS Economics & Planning

11 GDP (Gross Domestic Product) refers to Australia, GSP (Gross State Product) refers to a State, while GCP (Gross City Product) refers to a city. But for simplicity’s sake in this paper all different measures are referred to as GDP.

Confidential Copy 119

Labour productivity by industry (small area level)

For Melbourne, the labour productivity estimates have been disaggregated by industry for each Statistical Local Area (SLA), using the principles set out in the Australian System of National Accounts (ABS, 2000).

Two separate methods were used for capital intensive industries (Manufacturing, Wholesale trade and Transport, postal & warehousing) and labour intensive industries (all other industries except for Electricity, gas, water & waste services and Information, media & telecommunications).

For capital intensive industries, gross value added (GVA) per employee using the two digit Australia New Zealand Standard Industry Classification (ANZSIC 2006) level data sourced from Australian Industry, 2009-10 (cat. no. 8155.0) has been combined with detailed employment estimates for each SLA to calculate the GVA share of each SLA. This share has then been used to allocate the Melbourne total industry GVA to each SLA.

For labour intensive industries a quality adjusted labour input method was used. That is, average industry wage rates were estimated for each SLA and combined with total hours worked for each industry for each SLA. This provides a proxy for total factor income for the SLA. This data on average industry wage rates and hours worked can be sourced from the ABS Census. The SLA share was then used to allocate the Melbourne total industry GVA to each SLA.

Table 39 shows that the highest labour productivity SLAs are clustered around central Melbourne, the industrial zones in the south east, and the airport in the north. Much of the variation can be attributed to industry mix within each SLA. That is, the higher labour productivity service based industries tend to cluster around the CBD. Figure 73 presents such variation for the Professional, scientific & technical services industry. This map shows that a worker located in the central area of Melbourne has a higher labour productivity than a worker in the same industry located on the fringe of Melbourne.

Confidential Copy 120

TABLE 39. SELECTED SLA LABOUR PRODUCTIVITY SLA 1996 2001 2006 2011

Melbourne - Inner 55.2 73.2 84.2 85.3 Melbourne - S'bank-D'lands 42.9 44.7 72.8 105.0 Melbourne - Remainder 59.8 63.6 55.2 62.3 Port Phillip - St Kilda 40.7 50.3 56.8 57.9 Port Phillip - West 45.2 54.9 62.8 64.7 Stonnington - Prahran 47.6 53.2 52.9 53.0 Yarra - North 45.5 45.8 47.7 49.3 Brimbank - Sunshine 31.8 37.2 45.6 44.6 Maribyrnong 43.7 44.0 42.8 42.9 Moonee Valley - Essendon 40.7 44.5 45.2 44.5 Wyndham - North 23.5 32.9 45.6 45.4 Moreland - Brunswick 45.8 42.4 40.7 41.8 Banyule - North 34.4 38.2 40.7 40.1 Darebin - Preston 46.3 41.4 43.3 43.7 Hume - Broadmeadows 35.3 40.0 46.3 47.3 Whittlesea - South-West 29.8 33.8 42.8 43.3 Boroondara - Camberwell N. 42.8 47.7 52.9 53.8 Manningham - West 38.0 44.1 44.7 45.1 Monash - South-West 44.1 46.3 47.1 47.4 Whitehorse - Box Hill 40.8 46.1 49.3 48.3 Knox - South 32.3 40.6 46.9 48.5 Glen Eira - Caulfield 44.4 47.7 47.3 47.1 Gr. Dandenong - Dandenong 37.6 40.3 44.6 45.3 Frankston - East 28.7 42.3 40.9 40.1 Source: SGS Economics & Planning

Confidential Copy 121

FIGURE 73. SLA PROFESSIONAL, SCIENTIFIC & TECHNICAL SERVICES LABOUR PRODUCTIVITY

Source: SGS Economics & Planning

Accessibility model To estimate the induced change in households and employment from the road projects and investments into Box Hill, detailed analysis of changes in accessibility was undertaken by SGS. With the implementation of this project in the region two key components related to an areas access to jobs and services will be impacted and subsequently estimated. These are:  Changes in travel times; and/or  Changes in mode share uses (Public Transport versus Private Vehicles)

It should be noted that for this analysis the change in the distribution of people and firms is purely based on the induced effects of improved accessibility. While not captured in this analysis, in addition to these impacts, there can often be one or more targeted intervention developments near key locations (i.e. new stations) that may further capitalise on the improved accessibility of the corridor.

This model quantifies the observed relationship (at 1996, 2001 and 2006) that population and employment (by industry group) levels have with the relative accessibility (weighted travel time) of each SLA across Western Melbourne and the wider metropolitan area. In short, the model predicts where population and employment will gravitate given augmented accessibility in the regional rail network. This method is outlined in Figure 74 and in more detail following.

Confidential Copy 122

FIGURE 74. ACCESSIBILITY MODELLING APPROACH

Source: SGS Economics & Planning

Historical Relationship Analysis

Over the past 15 years there have been several major road infrastructure projects that have had a significant impact on the accessibility of certain locations across metropolitan Melbourne, such as:  1996 to 2001 – Western Ring Road  2001 to 2006 – Citylink  2006 to 2011 – Eastlink and major improvements to the Monash Freeway and the West Gate Bridge

Furthermore, over this period detailed information regarding the distribution and accessibility changes of Melbourne has been made available from the ABS Census and the DOT (Melbourne Integrated Transport Model). This data has all been aggregated to the Statistical Local Area (SLA) level for four time periods for this analysis (1996, 2001, 2006 and 2011).

While all these projects are road based projects they help to provide an evidence base for understanding the degree to which firms and households change their locational preferences as a result of shifts in the metropolitan accessibility contours.

A detailed statistical regression analysis was undertaken using this historical data to test and quantify this conceptual theory. This statistical analysis tests the following theory:

All else being equal, a SLAs share of metropolitan Melbourne’s total employment by industry sector is based on two broad factors:  Share of population/dwellings12  Accessibility13

In addition, if all else was equal, a SLAs share of metropolitan Melbourne’s total housing stock is based on two key factors:  the urban land supply14  Accessibility.

12 The share of population/housing has been included to capture the distributional changes of underlying population serving employment. 13 Where, accessibility is defined in its broadest sense. That is, a locations access to activity centres, skilled workers, employment opportunities, services, education, transport infrastructure, restaurants, etc. 14 The amount of urban land has been included to capture the varying geographic sizes of SLAs. That is, if SLA X is 10 times larger than SLA Y, all else being equal, it should have 10 times the amount of dwellings.

Confidential Copy 123

This concept suggests that the strong growth within a growth area is a result of increased land supply, while the recent strong growth within the inner city is a result of increased accessibility. Furthermore, if a location’s accessibility is improved it will increase the level of demand and in turn growth rate, similar to releasing land within a growth area. The strength of this relationship varies from industry to industry depending on its requirements and willingness/capacity to pay for more accessible locations.

For this analysis these two factors have been identified as the key considerations in the distribution of households and employment. However, it is noted that in reality the exact spatial distribution of households and employment is far more complex, particularly at a local level. There is a wide range of other localised issues, historical factors and development trends that also determine the locational choices of firms and people.

It has been assumed that all these other factors bearing out of localised issues are indeed inherent and therefore captured within the base case projections. Consequently, any changes observed under the Project Case scenario are over and above these other contextual issues.

In order to undertake this statistical analysis three key data variables were synthesised by SGS for each SLA across metropolitan Melbourne from 1996 to 2011:  SLAs share of total employment by industry sector and households  SLAs share of total urban land  SLAs relative accessibility

Population and employment

Historical employment and housing data for 1996, 2001, 2006 and 2011 by SLA has been collated from a range of ABS data sets including the past three Censuses (1996, 2001 and 2006), the ABS labour force survey data15 and ABS estimated resident population16.

Total employment by industry sector and households for each SLA is converted to a share of overall metropolitan Melbourne levels. The share of metropolitan Melbourne was used as it was assumed that there is a wide range of other external factors such as international and interstate migration and economic trends that influence the total amount of population and employment in a city as a whole. Furthermore, it is assumed that a project such as this will only have an influence in the reorganisation (or locational decisions) of people/firms within Melbourne and will not be able to influence a person from Sydney, for example. In other words, it was assumed that a project of this scale is unlikely to affect the overall employment outcomes for the MSD with only reorganisation of employment within the metropolitan economy envisaged.

Urban land

There is a range of data sources that could be utilised (such as the Victorian Planning Provisions) to calculate the amount of total urban land that an SLA has as a proportion of metropolitan Melbourne. However, many of these datasets are difficult to source consistently across a 15 year historical time period (1996-2011). Therefore, a geographic unit referred to as Urban Centre Locality (UCL) was used as a broad estimate of the amount of urban land. A UCL measures the broad extent of urbanisation for a city/town and is published by the ABS for the 1996, 2001 and 2006 Censuses. An estimate of the Melbourne UCL for 2011 was made by adding in all future 2010 Urban Development Program (UDP) broad hectare sites to the 2006 UCL. This provided a consistent measure across all time periods.

As the UCL grows for an area (i.e. due to green-field land being released) the proportion of total Melbourne UCL for other established locations was found to decrease. This statistic isolates dwelling growth related to changes in accessibility from growth related to increased development opportunities.

15 ABS Cat. No. 6202.0 16 ABS Cat. No 3201.0

Confidential Copy 124

Accessibility

SGS has developed a measure of accessibility within a specified geographical region and the ability to access overall economic activity across the wider MSD, known as Effective Job Density (EJD). The measurement of EJD has been outlined previously.

Quantifying changes to relative accessibility was performed through translating the absolute EJD’s across the four years of analysis into a 0 to 1 index, i.e. a relative EJD index. The index is created using the SLA that has the highest EJD. This was found to be Melbourne (C) – Inner consistently across all years of analysis, which provides a ceiling for the index (a score of 1). To provide a floor to the index the converse is performed. Consistently Yarra Ranges (C) – Central was found to have the lowest EJD and is thus given a score of 0.

The equation used to calculate the relative EJD ranking is shown below.

ܧܬܦ െ ܯ݅݊ܧܬܦ ൌ ௜ ܦܬܧݐ݅ݒ݈ܴ݁ܽ݁ ܦܬܧ݊݅ܯ െ ܦܬܧݔܽܯ ௜ Where:

ܧܬܦ௜= Effective Job Density for SLA i ܯ݅݊ܧܬܦௌ௅஺ = the SLA found to have the lowest EJD amongst all MSD SLAs ௌ௅஺= the SLA found to have the highest EJD amongst all MSD SLAsܦܬܧݔܽܯ

Agglomeration Model

This sub-section introduces the overall method (see Figure 75) for estimating and applying productivity elasticities with respect to agglomeration. The latter is measured by Effective Job Density (EJD). The EJD of an area is the sum employment in the area and the employment in all other areas divided by the travel time in reaching these external jobs, with travel time weighted for transport mode.

The first stage in the method is to assemble the four data sets required in this process. These are a travel time matrix, employment and population at a small area level, labour productivity by sector at a State and city level, and then at a small area level.

A travel time matrix shows how long it takes to travel from one small area in the city to all other small areas by car and public transport separately. This is produced within the transport modelling conducted externally to this process. Employment and population can be sourced from the ABS and other Government departments and is required for the EJD calculation. Labour productivity is released by the Australian Bureau of Statistics (ABS) at the State level which can then be broken down into a city and balance of State estimate. This is done via the same method the ABS uses in its National Accounting Framework. The city level labour productivity is then broken down into a statistical local area (SLA) level using the same principles set out by the Australian System of National Accounts.

The next stage is to calculate EJD at a small area level under a base case scenario. The two inputs are the travel time matrix and employment at a small area level. Once EJD has been calculated, a regression analysis using these estimates and labour productivity by sector can be conducted. This allows for the estimation of industry elasticities which show the impact on labour productivity of a doubling in EJD. The change in EJD brought about by a specific infrastructure or land use project can then be established in the same way the base case EJD is estimated. The inputs into this estimation will differ in terms of a changed travel time matrix or employment numbers.

Finally the impact on gross value added (GVA) arising from changes to labour productivity can be quantified. This is completed via the use of the industry elasticities applied to the estimated changes to

Confidential Copy 125

EJD to determine the changes to labour productivity. This altered labour productivity then flows through to a change in gross value added across industries and SLAs.

FIGURE 75. AGGLOMERATION METHOD – LABOUR PRODUCTIVITY

Labour Productivity by Sector (State and City Level)

Employment by Labour Productivity Travel Time Matrix Sector & Population by Sector (Small Area Level) (Small Area Level) (Small Area)

Calculate Effective Job Density (EJD) (Small Area Level)

Estimate Productivity Elasticity Versus EJD by Sector

Change in EJD Associated with Infrastructure or Land Use Initiative

Estimated Net Increase in GVA

Source: SGS Economics & Planning

Travel Time Matrix (Small Area Level)

A travel time matrix shows how long it takes to travel from one zone in the city to all other zones by car and public transport separately. They are also produced for a specific time period, such as morning or afternoon peak and off-peak. For the purposes of EJD calculation, SGS has used morning peak matrices.

In general, the headings going horizontally across the top of the matrix indicate the destination zone, and the first column going vertically down the left hand side of the matrix is the origin zone. An example is presented in Table 40, showing that if one is travelling from Melbourne – Inner to Melbourne – Remainder it will take 3.8 minutes on average to get there by car.

TABLE 40. EXAMPLE TRAVEL TIME MATRIX FOR PRIVATE VEHICLE, MINUTES

Destination Zone

Melbourne - Melbourne - Melbourne - Port Phillip - Port Phillip - Origin Zone Southbank- Inner Remainder St Kilda West Docklands Melbourne - Inner 1.3 3.8 2.9 7.1 4.1 Melbourne - Remainder 3.9 5.1 5.4 9.4 6.7 Melbourne - Southbank-Docklands 3.0 5.2 2.5 6.7 3.4 Port Phillip - St Kilda 8.1 9.7 7.6 3.0 6.3 Port Phillip - West 5.0 7.0 4.0 5.7 3.2 Source: SGS Economics & Planning

Also required with the travel time matrix is an estimate of the share of public transport use by workers travelling to their place of work. Including this share into the measure of effective job density enables a more ‘real life’ representation of the proximity (in terms of travel time) component of agglomeration. By

Confidential Copy 126

way of illustration, 68% of people working in the CBD of Melbourne travel to work on public transport, implying that the proximity to those jobs is closely related to public transport travel times. The other extreme can be seen in locations such as Cranbourne (an outer suburb of Melbourne), where 98% of workers travel to work using private vehicles17.

Throughout the estimation process, a small area has been defined as a statistical local area (SLA). However, as noted, a travel time matrix is generally produced at an even smaller geography due to the detailed nature of transport modelling. SGS has developed a methodology to convert these matrices from the travel zone level to an SLA level. This method could be applied to any travel time matrix given a concordance between the travel zones and SLAs. The steps involved in this method are outlined below, along with a simple worked example presented in Figure 76.

Step 1: Convert the travel zone by travel zone travel time matrix (or generalised cost matrix) into a SLA by travel zone travel time matrix.

௜൱ݎݐ݋ܽݎ݁݊݁ܩ ௜ ൊ ෍ݎݐ݋ܽݎ݁݊݁ܩ௑௝ ൌ෍ܶܶ௜௝ ൈ ൭ܶܶ ௜ఢ௑ ௜ఢ௑ Step 2: Convert the SLA by travel zone travel time matrix into a SLA by SLA travel time (or generalised cost) matrix.

௝ቍݎݐ݋ܿܽݎݐݐܣ ௝ ൊ ෍ݎݐ݋ܿܽݎݐݐܣ௑௒ ൌ෍ܶܶ௜௝ ൈ ቌܶܶ ௝ఢ௒ ௝ఢ௒ Where:

ܶܶ௜௝ = Travel Time from travel zone i to travel zone j ܶܶ௑௒ = Travel Time from SLA X to SLA Y ௜ = Population in Travel Zone iݎݐ݋ܽݎ݁݊݁ܩ ௝ = Employment in Travel Zone jݎݐ݋ܿܽݎݐݐܣ

It is possible the aggregation calculation from travel zone to SLA could be conducted in a number of ways. However, all methods should produce broadly the same result.

17 This method excludes travel times from other modes (bicycle or walk.)

Confidential Copy 127

FIGURE 76. WORKED EXAMPLE OF CONVERTING A TRAVEL TIME MATRIX TO SLA MATRIX

Source: SGS Economics & Planning

Calculate Effective Job Density (small area level) In calculating EJD SGS has used the level of employment relative to the time taken to gain access to that employment and the mode split that is currently experienced by those workers in their travel to employment. The formula used to calculate EJD at an SLA level is presented below.

ܲܶܯ݋݄݀݁ܵܽݎ݁௝ ൈܧ݉݌௝ ሺͳ െ ܲܶܯ݋݄݀݁ܵܽݎ݁௝ሻൈܧ݉݌௝ ܧܬܦ௜ ൌ෍ቆ ൅ ቇ ݒ݈݁ܶ݅݉݁௜௝ܽݎܸܶܲ ݒ݈݁ܶ݅݉݁௜௝ܽݎܶܶܲ ௝

Where:

ܧܬܦ௜= Effective Job Density for SLA i ܲܶܯ݋݄݀݁ܵܽݎ݁௝ = per cent of work trips which involve public transport for SLA j ܧ݉݌௝= number of jobs/employment within SLA j ݒ݈݁ܶ݅݉݁௜௝ = time it takes to travel on public transport from SLA i to SLA jܽݎܶܶܲ ݒ݈݁ܶ݅݉݁௜௝ = time it takes to travel by private vehicle from SLA i to SLA jܽݎܸܶܲ

The public transport mode share and public transport and private vehicle travel times will come from an external source, as outlined previously.

A worked example of how to calculate EJD is presented in Figure 64. Additionally, estimates for EJD in Melbourne are presented graphically on the map in Figure 68.

Confidential Copy 128

FIGURE 77. WORKED EXAMPLE OF EJD CALCULATION

Source: SGS Economics & Planning

FIGURE 78. EJD IN MELBOURNE

Source: SGS Economics & Planning

Estimate productivity elasticity versus EJD by sector

To measure agglomeration impacts, variations in labour productivity across the city for a particular industry should be observable. This relationship between labour productivity and EJD can be estimated via regression analysis. In this Melbourne illustration, this analysis was conducted using a translog formulation where the natural log of labour productivity levels for the respective industry is regressed against the natural log of effective job density by SLA.

Confidential Copy 129

Due to their small size within the metropolitan economy, both the Agricultural and Mining industries have been excluded from the analysis. Information, media & telecommunications and Electricity, gas, water & waste services industries have also been excluded due to the inability to effectively measure labour productivity at an SLA level. In both of these industries the vast bulk of GVA is attributed to the capital infrastructure covering the city. The labour productivity measure would allocate the GVA to the spatial clusters of workers in these industries, which would be misleading.

This step requires the use of three inputs, SLA EJD, SLA labour productivity by sector and a weighting based on SLA GVA for each industry. The relationship was estimated as follows:

Ž‘‰୬ሺ ୧ሻൌȾ଴ ൅Ⱦଵ ™Ž‘‰୬ሺ ୧ሻ൅ɂ୧

Where:

୧is the Industry Labour Productivity for SLA i  ୧ is the Effective Job Density for SLA i ™ is the weighted based on total SLA i industry GVA and ɂ୧ is a zero mean random disturbance

This analysis enabled the estimation of industry elasticities which allow the coefficients of the regression equation to be easily understood. These determine the degree to which changes in agglomeration affect labour productivity across a metropolitan region. The elasticity was estimated as the impact of doubling effective job density on labour productivity in each industry using the regression coefficients for Ⱦ଴ and Ⱦଵ.

Figure 79 illustrates the observed positive linear relationship between labour productivity and EJD for the Professional services industry in Melbourne as estimated by the equation above. The observations to the top right of the chart represent locations in Inner Melbourne.

The EJD elasticities for each industry in Melbourne are presented in Figure 80. The weighted (by industry GVA) total for all industries included in the analysis is 0.07. This can be interpreted as a doubling of EJD will result in an 7% increase in labour productivity in an area, on average.

Confidential Copy 130

FIGURE 79. SCATTER PLOT LABOUR PRODUCTIVITY AND EJD, PROFESSIONAL SERVICES

Source: SGS Economics & Planning

FIGURE 80. INDUSTRY EJD ELASTICITIES

Source: SGS Economics & Planning

Change in EJD associated with infrastructure or land use initiative

The change in EJD associated with an infrastructure or land use initiative is estimated via the same method used to calculate the base EJD as outlined previously. The difference between the base case EJD

Confidential Copy 131

and a project case EJD will depend on the project inputs and data availability. Both projected travel times and the share of public transport use could be affected by the project. Additionally the level of future employment in a particular location could change. A project case EJD should then be estimated using these altered inputs and compared to the base case EJD. A project case EJD can be estimated for several future time periods, such as 2016, 2021 and 2031. These should be compared to a base EJD estimated for the corresponding year.

Each of the changes to EJD inputs will have a differing impact on EJD. Increasing employment in a particular location in general has the largest impact on EJD. This is most likely to occur under a land use project which increases employment at a targeted location. A reduction in travel times (both public and private transport) has a significant impact on increasing EJD, but to a lesser extent than employment changes. Increasing the share of public transport use can have adverse impacts on EJD depending on the project. If a greater share of commuters are travelling on public transport which still has a longer average travel time than private vehicle, EJD will be made worse.

To examine the geographic distribution of changes to EJD from a project, the map in Figure 82 presents the percentage change in SLA EJD across metropolitan Melbourne for a given transport infrastructure project.

FIGURE 81. MELBOURNE EXAMPLE PERCENTAGE CHANGE IN EJD

Source: SGS Economics & Planning

Estimated Net Increase in GVA

To estimate the net increase in GVA arising from changes to EJD three broad steps are required. First the uplift in industry labour productivity must be estimated using the regression equation coefficients and percentage change in EJD between the base and project cases. Then, the estimated benefit is calculated by applying the changes in labour productivity to a fixed hours worked by sector to calculate gross value

Confidential Copy 132

added by sector. Also in this step, industry labour productivity growth rates for the metropolitan region are applied to alter the assumption of fixed level of labour productivity into the future. Finally, the changes to GVA by SLA and industry are summed to produce the total economic benefit of the project. These steps are outlined in more detail below.

Step 1: Estimate uplift in labour productivity using EJD and industry elasticities.

ܮܲ௝௜ ൌܮܲ௝௜ ൈቀሺܧܬܦ௜െܧܬܦ௜ሻ ൈ൫ͳ൅ܧ௝ ൯ቁ Where:

ܮܲ௝௜ = Labour Productivity with the Project for each SLA i and Industry j ܮܲ௝௜= Base Case Labour Productivity for each SLA i and Industry j ܧܬܦ௜ = Base Case Effective Job Density for SLA i ܧܬܦ௜ = Effective Job Density with the Project for SLA i ୨= Estimated Elasticity for each Industry j

Step 2: Estimate the economic benefits for each industry and SLA using the uplift in labour productivity.

ܧܤ௝௜ ൌ൫ܮܲ௝௜ ൈܪܴܵ௝௜൯െ൫ܮܲ௝௜ ൈܪܴܵ௝௜൯ Where:

ܧܤ௝௜ = Economic Benefit with Project for each SLA i and Industry j ܮܲ௝௜ = Labour Productivity with the Project for each SLA i and Industry j ܮܲ௝௜ = Base Case Labour Productivity for each SLA i and Industry j ܪܴܵ௝௜ = Number of Hours Worked for each SLA i and Industry j

Step 3: Estimate the total economic benefit of the project

ܧܤ ൌ ෍ ܧܤ௝௜

The results can be presented in a variety of ways depending on the requirements of the end user. SGS has presented the results at an aggregate metropolitan level under different labour productivity growth rates in a particular year. This allows for sensitivity testing of the final benefit. The results can also be broken down to an industry level for the metropolitan region showing which sectors gain the largest benefit from the project. The total economic benefit can then be split up into SLA benefits showing which geographic regions are likely to benefit the most. In some cases, such as cost benefit analyses, it may be necessary to estimate a yearly stream of benefits from the project. This can be done by interpolating the benefit from the start of the project, between the future years to the end of the project life. This also allows for the estimation of the net present value (NPV) of a project.

A worked example with illustrative numbers is presented in Figure 82 below.

Confidential Copy 133

FIGURE 82. WORKED EXAMPLE OF LABOUR PRODUCTIVITY IMPACT ON GVA

Regression EJD 2031 Base Project Uplift Retail Trade Health Care Education Results Geelong 174 175 175 – 174 = 1 Industry Elasticity 0.16 0.13 0.09

Ballarat 110 111 111 – 110 = 1

Labour Hours Worked Retail Trade Health Care Education Retail Trade Health Care Education Productivity: Base 2031 Input Data Geelong 22 32 27 Geelong 19,600,000 21,600,000 13,800,000 Ballarat 20 30 25 Ballarat 9,952,000 17,500,000 9,200,000

Step 1: Labour Labour Retail Trade Health Care Education Productivity: Retail Trade Health Care Education Productivity: Uplift Project Geelong 26 – 22 = 4 36 – 32 = 4 29 – 27 = 2 Labour Productivity Base * EJD Uplift * (1 + Industry Elasticity) Ballarat 23 – 20 = 3 34 – 30 = 4 27 – 25 = 2 22 * 1 * 32 * 1 * 27 * 1 * Geelong (1+0.16) = 26 (1+0.13) = 36 (1+0.09) = 29

20 * 1 * 30 * 1 * 25 * 1 * Ballarat (1+0.16) = 23 (1+0.13) = 34 (1+0.09) = 27

Gross Value Added = Labour Productivity * Hours Worked Step 2: Step 3: Calculations Gross Value Added Uplift Retail Trade Health Care Education Total

Labour Productivity Uplift * Hours Worked 2031 Sum of Industry Uplift

4 * 19,600,000 4 * 21,600,000 2 * 13,800,000 Geelong 72,359,362 = 33,413,593 = 34,299,877 = 4,645,892

3 * 9,952,000 4 * 17,500,000 2 * 9,200,000 Ballarat 89,752,784 = 25,177,898 = 51,115,648 = 13,459,238

Source: SGS Economics & Planning

Human capital benefits estimation

This section details the steps required to calculate the benefit to human capital arising from changes to EJD. Figure 83 presents an overview of the input data and steps required in this process. Similar inputs to the labour productivity method are required which have been outlined previously. These include estimating employment and population at a small area level, generating a base and project case EJD from travel time matrices and establishing the change in EJD from the project initiative.

The higher metropolitan level data on human capital stock must first be estimated and then broken down to a small area level. This small area level human capital stock data by age, sex and qualification level is combined with EJD in a regression analysis. Elasticities can then be estimated to determine the effect of a change in EJD on human capital stock. The net increase in human capital is then estimated using these three inputs.

Confidential Copy 134

FIGURE 83. AGGLOMERATION METHOD – HUMAN CAPITAL

Human Capital Employment by Stock by Age & Travel Time Matrix Sector & Population Qualification (Small Area Level) (Small Area Level) (Small Area Level)

Calculate Effective Job Density (EJD) (Small Area Level)

Estimate Human Capital Elasticity Versus EJD by Age & Qualification Group

Change in EJD Associated with Infrastructure or Land Use Initiative

Estimated Net Increase in Human Capital

Source: SGS Economics & Planning

Human Capital Stock by Age & Qualification (Small Area Level)

Human capital for the broader metropolitan region is calculated using the methodology outlined by the ABS, in Measuring Human Capital Flows for Australia: A Lifetime Labour Income Approach (cat. no. 1351.0.55.023). Melbourne estimates by sex and qualification level are presented in Table 41.

Using the same Lifetime Labour Income approach human capital estimates for each SLA in a metropolitan region can be produced. These estimates should be benchmarked to a metropolitan total to ensure consistency with the city, State and Australian estimates. The data at the SLA level can be sourced from the ABS Census which releases information on individual income broken down by age, sex and qualification level. As an example, Figure 84 presents the variation in life time labour income for men with a bachelor degree for four SLA’s in Melbourne. SLAs with higher EJD have higher life time labour income.

Confidential Copy 135

TABLE 41. ESTIMATE OF MELBOURNE’S HUMAN CAPITAL ($ BILLIONS)18 1996 2001 2006

Men Higher Degree 26.3 35.0 54.8 Bachelor Degree 117.2 154.0 201.0 Skilled Labour 153.9 187.7 213.3 Unqualified 222.8 243.2 346.4 Total 520.2 619.9 815.4 Women Higher Degree 10.9 19.3 37.9 Bachelor Degree 86.8 130.7 182.5 Skilled Labour 64.9 80.3 103.1 Unqualified 199.6 219.7 261.3 Total 362.2 450.0 584.8 Total Higher Degree 37.1 54.3 92.7 Bachelor Degree 204.1 284.6 383.5 Skilled Labour 218.8 268.0 316.3 Unqualified 422.4 462.9 607.7 Total 882.4 1,069.8 1,400.3 Source: SGS Economics & Planning

FIGURE 84. GROSS ANNUAL INCOME PER CAPITA, MALE BACHELOR DEGREE

Source: SGS Economics & Planning

18 To maintain consistency with the 1351.0.55.023 publication all estimates are measured in 2001 constant dollars.

Confidential Copy 136

Estimate human capital elasticity versus EJD by age & qualification group

Using regression analysis, the relationship between EJD and human capital group for each SLA has been established. To maintain consistency with the labour productivity analysis the same functional form of equation is used. Separate regressions have been done for each education level / age / sex grouping. As well as the EJD variable a Socio Economic Index For Areas (SEIFA) variable has been included. The index of Economic Resources has been included to account for the level of income and assets which are available to households for initial investments in education. This data is produced by the ABS and is freely available. Elasticities can then be calculated to estimate the impact of EJD on human capital by qualification, age and sex type.

The relationship can be estimated as follows:

୕୅ୗ Ž‘‰୬ሺ ୧ ሻൌȾ଴ ൅Ⱦଵ ™Ž‘‰୬ሺ ୧ሻ൅Ⱦଶ ୧ ൅ɂ୧

Where: ୕୅ୗ ୧ is the Human Capital for SLA I, qualification level Q, age group A and sex group S  ୧ is the Effective Job Density for SLA i  ୧ is the SEIFA index ranking for SLA i ™ is the weighted based on total SLA i population, qualification level, age group and sex group ɂ୧ is a zero mean random disturbance

The map presented in Figure 85 shows the spatial variation in Bachelor Degree Human Capital across metropolitan Melbourne for males aged between 40 and 44. The data has been split into quintiles, with the 5th quintile being the largest stock of Human Capital, and the 1st the lowest level of stock. The highest level of human capital for males with a bachelor degree is concentrated around the central core of Melbourne and towards the eastern suburbs.

Confidential Copy 137

FIGURE 85. BACHELOR DEGREE HUMAN CAPITAL STOCK, MALES AGED 40 TO 44

Source: SGS Economics & Planning

Estimated Net Increase in Human Capital

To estimate the net increase in GVA from human capital impacts arising from changes to EJD three broad steps are required, similar to the labour productivity method outlined previously. First the uplift in human capital by qualification, age and sex grouping must be estimated using the regression equation coefficients and percentage change in EJD between base and project case. Then, the estimated change to human capital is calculated by applying the changes in human capital stock to a fixed population projection by qualification, age and sex grouping. This then allows the calculation of the total human capital uplift across the metropolitan area. Finally, the changes to human capital by SLA and qualification, age and sex grouping are summed. This is then multiplied by a capital services rate which assumes an elasticity between the impact of changes to human capital on GVA. These steps are outlined in more detail below.

Step 1: Estimate uplift in human capital using EJD and elasticities.

ொ஺ௌ ொ஺ௌ ܳܣܵ ܪܥ௜ ൌܪܥ௜ ൈ ൫ሺܧܬܦ௜െܧܬܦ௜ሻ ൈ ሺͳ൅ܧ ሻ൯ Where: ொ஺ௌ ܪܥ௜ = Human Capital with the Project for each SLA i and qualification level Q, age group A and sex group S ொ஺ௌ ܪܥ௜ = Base Case Human Capital for each SLA i and qualification level Q, age group A and sex group S ܧܬܦ௜ = Base Case Effective Job Density for SLA i ܧܬܦ௜ = Effective Job Density with the Project for SLA i ܧொ஺ௌ = Estimated Elasticity for each human capital qualification, age and sex grouping

Confidential Copy 138

Step 2: Estimate the total human capital benefits for each qualification, age and sex grouping and SLA using the uplift in human capital.

ொ஺ௌ ொ஺ௌ ொ஺ௌ ொ஺ௌ ொ஺ௌ ܶܪܥܤ௜ ൌ൫ܪܥ௜ ൈ ܲ݋݌௜ ൯ െ൫ܪܥ௜ ൈ ܲ݋݌௜ ൯ Where: ொ஺ௌ ܶܪܥܤ௜ = Total Human Capital Benefit for each SLA i and qualification level Q, age group A and sex group S ொ஺ௌ ܪܥ௜ = Human Capital with the Project for each SLA i and qualification level Q, age group A and sex group S ொ஺ௌ ܪܥ௜ = Base Case Human Capital for each SLA i and qualification level Q, age group A and sex group S ொ஺ௌ ܲ݋݌௜ = Population for each SLA i and qualification level Q, age group A and sex group S

Step 3: Estimate the total benefit to GVA of the project

ொ஺ௌ ܧܤ ൌ  ෍ ܶܪܥܤ௜ ൈܧ Where: ܧܤ= Economic Benefit to GVA ܧ = elasticity of impact of human capital on GVA (9.7%)

The results can be presented in a variety of ways. SGS has presented the results at an aggregate metropolitan level under the four main qualification levels (Unqualified, Skilled Labour, Bachelor Degree and Higher Degree). The results can also be broken down to male and female by qualification level. The total economic benefit can then be split up into SLA benefits showing which geographic regions are likely to benefit the most. In some cases, such as cost benefit analyses, it may be necessary to estimate a yearly stream of benefits from the project. This can be done by interpolating the benefit from the start of the project, between the future years to the end of the project life. This also allows for the estimation of the net present value (NPV) of a project.

A worked example with illustrative numbers is presented in Figure 86 below.

Confidential Copy 139

FIGURE 86. WORKED EXAMPLE OF HUMAN CAPITAL BENEFIT

Regression Skilled Bachelor Higher Unqualified Results Labour Degree Degree Qualification Type 0.09 0.09 0.09 0.09 Elasticity

Skilled Bachelor Higher Population Unqualified Total Labour Degree Degree Geelong 60,000 48,000 23,500 3,200 134,700

Ballarat 40,000 22,000 15,000 1,500 78,500 Input Data Human Capital: Skilled Bachelor Higher Unqualified Total Base Labour Degree Degree Geelong 433,000 538,000 747,000 600,000 2,318,000

Ballarat 192,000 238,000 301,000 298,000 1,029,000

Step 1: Human Capital: Skilled Bachelor Higher Unqualified Total Project Labour Degree Degree Human Capital Base * EJD Uplift * (1 + Qualification Type Elasticity)

Geelong 450,830 560,153 777,759 624,706 2,413,448

Ballarat 207,615 257,356 325,480 322,236 1,112,688

Human Capital: Skilled Bachelor Higher Unqualified Total Uplift Labour Degree Degree Geelong 17,830 22,153 30,759 24,706 95,448

Calculations Ballarat 15,615 19,356 24,480 24,236 83,688

Step 2: Step 3: Total Uplift Unqualified Skilled Labour Bachelor Degree Higher Degree Total

Human Capital Uplift * Population * 9.7% Sum

17,830 * 60,000 * 22,153 * 48,000 * 30,759 * 23,500 * 24,706 * 3,200 * Geelong 0.097 0.097 0.097 0.097 284,698,951 = 103,768,721 = 103,145,629 = 70,115,782 = 7,668,820 15,615 * 40,000 * 19,356 * 22,000 * 24,480 * 15,000 * 24,236 * 1,500 * Ballarat 0.097 0.097 0.097 0.097 141,038,933 = 60,587,258 = 41,306,625 = 35,618,681 = 3,526,368 Source: SGS Economics & Planning

Confidential Copy 140

REFERENCES

City of Melbourne (1985) Strategy plan 1985, adopted by council 14 May 1985

Department of Planning and Community Development (2008) Melbourne @ 5 Million, Melbourne, Australia

Department of Planning and Community Development (2012) Melbourne’s Strategic Planning History, access 12 June 2012 at:

Eddington, R. (2008) ‘Investing in Transport: East West Link Needs Assessment’

Freestone, R. (2008) The City Past and Present, eMelbourne, published by School of Historical Studies Department of History, The University of Melbourne, access 12 June 2012 at:

Graham, D.J. (2006) ‘Agglomeration, Productivity and Transport Investment’ in Journal of Transport Economics, Vol 41. No. 3.

Infrastructure Partnerships Australia (2006). Case Studies -CityLink Melbourne, Melbourne.

Institute of Engineers Australia (2002) Journey and arrival: the story of the Melbourne CityLink, Institute of Engineers Australia, Victoria

Marshall, K. (2011) The Age, “I was there. City Loop inauguration”, 11/25/2011

Melbourne City Council (1970) Facts on the financing of the proposed underground rail loop for Melbourne, November 20, 1970

Metropolitan Melbourne Board of Works (1954) Melbourne Metropolitan Planning Scheme 1954, Melbourne, Australia

Metropolitan Melbourne Board of Works (1971) Planning Policies for the Melbourne Metropolitan Region, Melbourne, Australia

Metropolitan Town Planning Commission (1929) Plan of General Development, Melbourne, Australia

Morris (2006) ‘Australia: Equity, Imbalance and Egalitarianism; in Blindenbacher, R. and Karos, A.O. (eds) Dialogues in the practice of fiscal federalism: comparative perspectives.

Muhammad, I. And Low, N. (2006) Mega projects in transport and development: background in Australian case studies: City Link motorway expansion, Melbourne, GAMUT, Melbourne

Perrott, Lyon, Timlock, Kesa, and Connell (1975) Box Hill transport centre: a Ministry of Transport project, Melbourne, Victoria

Public Transport Victoria (PTV) Corporation (2012) City Loop History, accessed 23rd April 2012, from:

Confidential Copy 141

Rawnsley, T.; Finney, B and Szafraniec, J. (2011) Riding the Smart Bus: Knowledge Workers, Agglomeration Economies and Public Transport Us, Presented to the AITPM National Conference, Melbourne.

Rawnsley, T.; and Spiller, M. (2010) Melbourne's Transformation: Rust Belt to Renaissance, Presented to the Knowledge Cities World Summit, Melbourne.

Rawnsley, T.; and Szafraniec, J. (2010) Agglomeration and Labour Productivity in Australian Cities, Presented to the Knowledge Cities World Summit, Melbourne.

Rogan, E. (2012) The Melbourne city underground rail loop – a case study in capital non-programming, Australian Planner, 49:1, 20-26

Rumpf, P. And Kiss, R. (1994) ‘Public policy, a major project and the role of engineers’, 1994 International conference on engineering management, Melbourne 10-12 April, 1994, The Institution of Engineers Australia

Russell, B (2000). Audit Review of Government Contracts, Contracting, Privatisation, Probity and Disclosure in Victoria 1992 -1999, An Independent Report to Government (Vol. 2 –Case Studies). Melbourne: State Government of Victoria

SGS Economics and Planning (2011) Funding Growth Areas in a National Urban Policy – Background Paper, prepared for National Growth Alliance.

VicRoads (1994) Western Bypass & Southern Bypass: Environment Effects Statement, Supplement (No. ANB95/04). Melbourne.

VicRoads (1994a) Linking Melbourne, Victorian Government

VicRoads (2008b). CityLink Project Overview. Retrieved 2 January 2008 from www.vicroads.vic.gov.au/Home/RoadsAndProjects/RoadProjects/InnerCity/CityLink/ProjectOverview.ht m

Victorian Government (1995) Living suburbs: a policy for metropolitan Melbourne into the 21st century, Victoria, Australia

Victorian Government and the Melbourne City Council (1994) Victoria Capital City Policy, Victoria, Australia

Warburton, D. and Hendy, P. (2006) International Comparison of Australia’s Taxes, Australian Treasury.

Confidential Copy 142 www.sgsep.com.au

Contact us

BRISBANE PO Box 117 Level 1, 76 McLachlan Street Fortitude Valley QLD 4006 +61 7 3124 9026 [email protected]

CANBERRA Level 1, 55 Woolley Street Dickson ACT 2602 +61 2 6262 7603 [email protected]

HOBART Unit 2, 5 King Street Bellerive TAS 7018 +61 (0)439 941 934 [email protected]

MELBOURNE Level 5, 171 La Trobe Street Melbourne VIC 3000 +61 3 8616 0331 [email protected]

SYDNEY Suite 12, 50 Reservoir Street Surry Hills NSW 2010 +61 2 8307 0121 [email protected]