Adequacy of the Auckland Region s Residential Land Supply

Transcription

1 Adequacy of the Auckland Region s Supply Prepared for September 2008 Status: Final Report

2 Table of Contents EXECUTIVE SUMMARY INTRODUCTION Research Background Objectives Methodology Report Structure HOW DOES THE LAND SUPPLY PROCESS WORK? Aims and Objectives Regulatory Processes Governing Land Supply in the Auckland Region Land Supply Process within Auckland Region AUCKLAND REGION RESIDENTIAL LAND SUPPLY Auckland Region Capacity for Growth Study 2008 (Theoretical Supply) Summary of Updating Capacity Process, Residential Capacity, Analysis of Supply Identification of Current Stage in the Consent/Development Process Model Months to Market Conclusions AUCKLAND REGION RESIDENTIAL LAND DEMAND, Housing Demand Model Housing Demand Model Methodology and Datasets Demand for Dwellings in Auckland Region, 2008 to ADEQUACY OF RESIDENTIAL LAND SUPPLY Reconcile Supply & Demand Figures by Time Period Allocation Approach Results: TA Summary Outcomes Results: Years till Exhaustion Results: Suburb Area Outcomes Analysis Limitations and Caveats RIGIDITIES AND BLOCKAGES General Limiting Urban Expansion Intensification...64

3 6.4 Landbanking Fragmentation Regulatory Processes Infrastructure Constraints CONCLUSIONS Supply of Residential Capacity Residential Capacity Distribution Rural land Capacity Infill Consumption and Redevelopment Opportunities Acceptance of Intensification Additional Research and Next Steps...77 Table of Maps Map 1.1: Auckland Region Local Authorities (MUL Outline Added in Purple)...7 Map 5.1: Auckland Regional Residential Capacity as at June 30, Map 5.2: Auckland Regional Residual Residential Capacity, Table of Figures Figure 1.1: Adequacy Framework Approach...11 Figure 5.1: Growth Allocation Process Summary...47 Table of Tables Table 3.1: Auckland Region Total Household Capacity under current policy Table 3.2 Auckland Residential Capacity Table 3.3 Auckland Residential Capacity by Density Type Table 3.4: High Redevelopment Scenario Residential Capacity Table 3.5: High Redevelopment Scenario Capacity by Density Type Table 3.6: Restricted Rural Scenario Residential Capacity Table 3.7: Restricted Rural Scenario Capacity by Density Type Table 3.8: No Rural Scenario Residential Capacity Table 3.9: No Rural Scenario Capacity by Density Type Table 3.10 Total Residential Capacity by Density Category and TA, Table 3.11 Current Stage in Consent/Development Process (Milestones) and Periods Between Stages...33 Table 3.12: Auckland Vacant by Consent Process Stage, Table 4.1 Estimated Dwellings in Auckland Region by Territorial Authority,

4 Table 4.2 Incremental Demand for Dwellings in Auckland Region by Territorial Authority, 2008 to 2031 (based on Medium Population Projection)...42 Table 4.3 Percentage Share of Growth in Demand for Dwellings in Auckland Region by Territorial Authority, 2008 to 2031 (based on Medium Population Projection)...42 Table 4.4 Incremental Demand for Dwellings in Auckland Region by Dwelling Type, 2008 to 2031 (based on Historical Intensification)...43 Table 4.5 Incremental Demand for Dwellings in Auckland Region by Dwelling Type, 2008 to 2031 (based on Melbourne Sydney Intensification)...43 Table 5.1: Auckland Region Residential Capacity (Units) by TA...44 Table 5.2: Auckland Region Medium Residential Household Growth by TA,...44 Table 5.3: Auckland Region Total Households (Medium Growth) by TA, Table 5.4: Household Migration by TA of Usual Residence and TA of Usual Residence 5 Years Ago, Table 5.5: Allocation of Households to Total Residential Capacity Table 5.6: Allocation of Households to Conventional Residential Capacity Table 5.7: Allocation of Households to Medium Residential Capacity Table 5.8: Allocation of Households to High Residential Capacity Table 5.9: Years till Total ARC Capacity Exhaustion by Density Type Table 5.10: Years till Capacity Exhaustion by Density Type: Base Run, high growth...55 Table 5.11: Years till Capacity Exhaustion by Density Type: Intensification, Medium Growth Table 5.12: Years till Capacity Exhaustion by Density Type: Restricted Rural, Medium Growth Table 5.13: Years till Capacity Exhaustion by Density Type: No Rural, Medium Growth...56 Table 5.14: Years till Capacity Exhaustion by Density Type: High Redevelopment, Medium Growth Table 6.1: Future Proposed Residential Areas Scale and Timing...63 Table 6.2: Multistage Greenfield Landholdings...68 Table 6.3: Vacant and Vacant Potential Land Holdings Without Consent or Development Activity by TA, Table 6.4: Nature of Infrastructure Constraints by TA,

5 Executive Summary The Department of Housing and Building wishes to establish whether there is an adequate supply of residential land in the Auckland Region to cater for future housing demand. The objective of the research is to determine the availability of land for housing and its adequacy to meet future demand, to identify rigidities and blockages that exist in the land supply process and the extent to which they impact on the delivery of land to market in an efficient manner especially in relation to higher density housing. Four key research questions have been posed: i. How does the land supply process work? In the Auckland Region land supply is governed by the overarching Growth Concept contained within the Auckland Regional Growth Strategy. This is based around the idea of a Compact City form, with growth focused in and around centres and along transport corridors, (primarily to support public transport) and limited expansion of the city at the borders. Effectively the city is constrained by the Metropolitan Urban Limit (MUL) that divides the region between urban land uses and rural uses and is designed to protect the regions natural and heritage resources. This occurs under the requirements laid out in the Local Government (Auckland) Amendment Act 2004 and the Regional Policy Statement. Through the Regional Growth Forum and the Memorandum of Understanding signed by the regions local authorities, the Growth Concept is implemented through the Sector Agreements (Northern, Central and Southern). Extensions to the MUL are governed by a framework within the RPS and are limited so as not to compromise intensification. In practice, extensions are complex, difficult and take approximately 5 years to work through the consent process. Within the confines of the MUL, the market forces of demand and supply apply. The MUL has the effect of limiting the total quantity of land supply, providing incentives for more efficient land use within urban areas. It also potentially creates opportunities for monopolistic behaviour by land owners, which could lead to inefficient allocation of the land, price increases and may delay the rate of release of residential land. This study did not find compelling evidence of land banking in the Auckland region. However, there is evidence that land supply is becoming increasingly constrained, which is likely to contribute to increased residential land prices throughout the city, potentially impacting on housing affordability in the region. While there appears to be general support for the Compact City growth objectives amongst stakeholders in the Auckland region, the Regional Council and Territorial Local Authorities face increasingly difficult and complex challenges in realising this goal. Adequate land supply implies a careful balancing act between constraining urban land Page 1

6 supply, through the application of the MUL or similar regulation, on the one hand and a measured approach to urban expansion on the other hand to avoid sharp spikes in land prices. As constraints on urban land supply become increasingly binding, there is potential for market power to accrue to a limited number of land owners potentially resulting in owners holding back land from development to maintain prices at high levels. ii. How much vacant land is available for residential housing, where is it located and who owns it? Vacant land represents a small portion of the total available residential capacity. At present vacant and vacant potential land has the capacity to house 25,270 units or 13.4% of the total. To this can be added capacity contained within Structure Plan areas (earmarked for future growth) of 34,200 units (18.1% of total capacity). The majority of this land is held reasonably tightly by major land owners and developers for a range of reasons (market power, convenience, to ensure sufficient ongoing development work, other historical reasons). Evidence of land banking is difficult to prove through this study process and needs additional work. At present (June 2008) and under current zoning regulations, development practices and market preferences, there is capacity for some 188,690 dwellings within the Auckland region. This includes significant capacity within the rural parts of the region that in reality may not be available for general urban growth due to market preferences and council policy response. Overall, 77% of capacity resides within the MUL however only 51% of conventional density capacity does (that is, capacity for development of standalone houses on 400sqm or more of land). This represents a significant risk to growth in the Auckland region as it implies that, unless market preferences towards intensive development change significantly or there is an increase in conventional density residential capacity (e.g. through extension(s) to the MUL), the Auckland region is unlikely to be able to accommodate projected future growth. If the rural capacity (30,570 units) is excluded (or only partially available) then the pressures on residential land supply within the MUL increase significantly. The second key element of supply is the high proportion of capacity that lies within business areas (40%). This is a second risk as it is both likely to be competed for strongly by other commercial uses and is generally high density (less than 200sqm land area per unit). At present, these environments have not favoured families with children due to limited amenity and limited housing choices. Currently around 12% of the region s households reside within business areas. This study shows this would need to change markedly in order for the region to achieve its Compact City goals. Should the regions metropolitan TAs focus on limiting infill and driving redevelopment, then capacity rises to around 199,000 more if site aggregation and comprehensive Page 2

7 redevelopment is facilitated. However, the reality of limiting or removing rural capacity from the mix sees capacity drop to between 168,000 and 178,000, depending on the level of restrictions imposed. iii. Is there an adequate supply of land suitable for residential housing development? The medium population growth scenario sees some 210,600 households added to the Auckland region between 2008 and Depending on the supply scenario chosen this represents a supply shortfall of between 12,000 and 43,000, with the most likely shortfall being around 22,000. Therefore, based on current zonings, the residential supply capacity will not meet the region s growth needs over the next 25 years. While cities do not necessarily plan to have 25 years of capacity, what is of concern is the composition of supply capacity. In Auckland, conventional density capacity is the traditional housing choice of most New Zealanders. This study shows that, under the Base Case scenario, conventional density housing is projected to be exhausted by In Auckland City, North Shore and Manukau, conventional density land supply is projected to be exhausted between 2015 and 2016, or in 7 to 8 years time. These figures, which include residential capacity on rural land, would be further reduced if this form of development is restricted; at least one TA is considering moves to limit this style of development. Under a high growth scenario, residential capacity runs out even earlier. This means that significant changes in households dwelling preferences towards more intensive living and/or an increased supply of Greenfield land, would be required in order to avoid shortfalls of Greenfield land in Auckland. Conventional economic thinking suggests that constrained land supply will raise prices and force alternate housing choices (including, potentially, outside of the region). The economic and social consequences of the supply limitations and changes to housing styles should be studied so that the right policy settings and choices can be made around future urban form. iv. Are there rigidities or blockages in the land supply process that may be contributing to an inefficient use, allocation and rate of release of land, particularly for the development of higher density housing? Research with land owners developers, planners and infrastructure providers has identified a range of rigidities and blockages within the land supply process. These can be summarised as; The limitation on urban expansion by the MUL. The MUL is, by design, contributing to a short supply of Greenfield land in urban areas. There is a lot of store placed on this regulatory tool as a means of achieving greater intensification but so far progress towards a Compact City future has been disappointing. As constraints on Page 3

8 urban land supply become increasingly binding, there is potential for market power to accrue to a limited number of land owners. As the process for expansion of the MUL is long and complex, it has effectively entrenched the current volume of land. In total only around 4,500 units have been added between 2006 and 2008 through MUL extensions. Private plan changes are seeking to add a further 5,000, however these figures combined add to around 1 year s worth of regional demand growth. Ineffective nature of intensification policies enacted through zone changes. The market has, with few exceptions, not taken up intensification opportunities on a large scale, widening the gap between intensification objectives and market demand for house and land type development. In short, conventional density housing on Greenfield land and general urban infill tends to be taken up at a greater rate than more intensive (more in tune with the objectives of a Compact City) options. Landbanking. Evidence indicates large blocks of greenfield land is reasonably tightly held within the MUL. However, even without concentrated ownership, there is an inadequate supply of currently zoned Greenfield land within the MUL to alleviate growth pressures. This is an area for further research. Fragmentation of land holdings especially in and around town centres makes it difficult to comprehensively redevelop sites to achieve a significant increase in residential densities accompanied by high amenity levels associated with ease of access to urban facilities and public transport. In a few locations, agencies such as Housing New Zealand have both the capacity and the concentrated holdings to achieve comprehensive redevelopment. However, it is likely that Councils or central government agencies would need to play a significant role in aggregating fragmented holdings if intensification is to be achieved. If there is to be land aggregation for achieving intensification objectives, agencies may need additional powers. The complexity, demands and volume of regulatory processes and the capabilities/capacities available to manage and run them. Skill shortages affect both the consent processing side and the application preparation side of the process and significantly impact on the ability to achieve the aim of a Compact City Page 4

9 1 Introduction 1.1 Research Background The Department of Housing and Building wishes to establish whether there is an adequate supply of residential land that is either ready or close to ready for development in the Auckland Region. The Department considers adequacy in a broad manner that looks at both the simple allocation of sufficient houses for the population anticipated, through to particular types of houses and dwelling stocks to match the housing preferences of the anticipated population. Market Economics (ME) and Harrison Grierson Consultants Limited (HG) have been engaged by the Department of Building and Housing (DBH) to undertake this research into the adequacy of Auckland Region s residential land supply. This report sets out the approaches and methodologies adopted and the findings of that research. 1.2 Objectives The objective of the research as set out in the project brief is to: survey land zoned residential (or otherwise available for residential development) in Auckland to determine its availability for housing and its adequacy to meet future housing demand. In doing so, the research will identify the rigidities or blockages (regulatory or non regulatory) that exist in the land supply process and the extent to which they may be causing land to be inefficiently used, allocated and released, particularly for the development of higher density housing.. Underlying this overarching objective are the following two key research aims: To identify how much land is available for residential housing in Auckland Region by location and by patterns of ownership. To identify whether there is an adequate supply of land suitable for residential housing development in the Auckland Region. The geographical coverage of this study is limited to the Auckland Regional Council (ARC) area. This is comprised of seven local authorities the North Shore, Waitakere, Auckland, Manukau City Councils and the Rodney and Papakura District Councils, and that portion of the Franklin District Council s territorial area that falls within the Auckland Region (Map 1.1). The project brief describes four key research questions which are to be addressed. These are as follows: Page 5

10 i. How does the land supply process work? The brief requires the description of an evidence based framework for analysing the land supply process, which identifies relevant incentives and drivers of behaviours of land owners and other relevant market participants. The framework is expected to be based on accepted theories and evidence relating to the operation of land supply in housing markets. The effect of the housing cycle on incentives and market dynamics is expected to be considered. The framework is required to distinguish between land at different stages of the development process (e.g. un subdivided, sub divided) and different types (e.g. single sections, infill). ii. How much vacant land is available for residential housing, where is it located and who owns it? The brief requires that the availability, location and patterns of ownership of land zoned residential to be identified. The analysis is required to identify land at different stages of development (e.g. ready or close to ready for development). The analysis is also required to take into account, but not be restricted to, actual potential supply of land, the permitted density of housing, infill and greenfield development. iii. Is there an adequate supply of land suitable for residential housing development? The brief requires an assessment to be made of the adequacy and suitability of residentialzoned land for housing and the likely level and timing of future demand. This will take into account the density of housing that zoning allows. The analysis will also consider factors such as the local geology (which may be relevant for cost of buildings, development risks and accessibility), relevant local and regional council regulations and infrastructure issues. iv. Are there rigidities or blockages in the land supply process that may be contributing to an inefficient use, allocation and rate of release of land, particularly for the development of higher density housing? The brief requires information to be collected from land owners and other relevant market participants on the issues, incentives and drivers that affect decisions about making land zoned residential available for housing development in practice. This includes considering issues of timing of supply decisions and the nature of the developments undertaken. The resulting information will be analysed to identify any unnecessary rigidities or blockages in the residential land supply process in practice, including: The role of local and regional council regulations Infrastructure issues Land ownership issues Other relevant factors. Page 6

11 Map 1.1: Auckland Region Local Authorities (MUL Outline Added in Purple) Source: Page 7

12 1.3 Methodology Framework Development In order to develop an assessment tool capable of answering the complex questions of adequacy of residential land supply, it was firstly necessary to understand the role land plays in the residential development process. This required establishing a firm theoretical foundation of the land supply process and how that plays out in the Auckland Regional environment. To achieve these aims we have: Carried out a literature review into the land supply process, the property and housing cycles and the Auckland residential market Developed a land supply process framework in draft form based on the collective knowledge of industry professionals Refined the assessment framework in consultation with local authorities, land owners and developers to reflect the realities of the development process in the Auckland Region as well as crosschecking with accepted land supply theory Demand and Supply Framework In order to achieve the key research objectives of assessing residential capacity and testing its adequacy, it is necessary to model the following key determinants of capacity and adequacy: The availability and location of land for development or redevelopment; The stage that this land is currently at in the development/consent process; The factors that would determine the land completing the development/consent process, and being available to the market in the form of either sections or units (i.e. sections which can be purchased on which to build homes, or otherwise completed houses, apartments available for purchase and occupation); The timing associated with completing this process; How the supply of sections/units interacts with the forecast demand over time across the Region, and by household type and type of residential accommodation. The flowchart in Figure 1.1 summarises the methodology that we have developed and utilised for this research project. In parallel with developing and undertaking this methodology, we have developed the spreadsheet based framework for modelling supply and demand both spatially and over time. Sections 5, 6 and 7 of this report describe each of the steps shown on the diagram. Page 8

13 For the interim dataset (covering the all larger landholdings) information was also collected on ownership of the majority of vacant and vacant potential land holdings within the Metropolitan Urban Limits (MUL). Assessments were also made in respect of the suitability of the land for development benchmarked against the development potential provided for by the relevant District Plan zoning rules covering each land parcel Interviews with Developers Research Question 4 is focused on the identification of rigidities and blockages in the land supply process that may be contributing to the inefficient use, allocation and rate of release of land for residential development. To answer this it was necessary to collect information from key industry operators. Interviews were undertaken with 12 major developers/landowners in the Auckland Region selected on the basis of scale, diversity of markets covered, distribution across the region and ownership. These interviews were undertaken on the basis that responses would be synthesised, rather than being specifically attributed to a particular interviewee/company. Between them, the group s land interests cover the following areas: Hobsonville Airbase and parts of the Babich Structure Plan area in Waitakere City The Long Bay Structure Plan area in North Shore City Parts of Orewa West structure plan area in Rodney District The Mt Wellington Quarry Structure Plan area in Auckland City Parts of the Flatbush Structure Plan area in Manukau City Large portions of the Hingaia and Takanini Structure Plan areas in Papakura District. The group also has, between them, smaller land interests elsewhere in the Region. These interviews were undertaken for three main purposes: 1. To assist in the development, refinement and calibration of our assessment framework and allocation model. 2. To provide us with advice in respect of intentions for specific parcels of land. 3. To provide us with qualitative information as to their drivers and decision making processes and their views on rigidities and blockages in the market. In addition to major land developers, interviews were held with Councils responsible for the delivery of key infrastructure and in charge of the land development and building consenting processes. Council staff was also able to identify the timing and scale of key Page 9

14 developments particularly on brownfield sites that were not owned by the developers interviewed. This provided an alternative insight into the development process, rigidities and blockages faced as well as detail on the timing of infrastructure development and the processes involved in matching infrastructure development with land use change. 1.4 Report Structure The remaining parts of this report are set out as follows: Section 2 describes how the land supply process works in the Auckland context and describes the framework for assessing land supply (Research Question 1). Section 3 describes the Auckland Regional Council Capacity for Growth (2008) data for residential land supply in 2006 and outlines the methodology for contemporising the estimates of land supply to 2008 (Research Question 2). Section 4 outlines anticipated household growth by dwelling type for the Auckland Region from 2008 to 2031 (Research Question 3). Section 5 compares household demand with residential supply and sets out findings in relation to adequacy of supply by area (Research Question 3). Section 6 identifies rigidities and blockages inherent in the land system, impacting on the delivery of appropriate residential land to the market (Research Question 4) Section 7 presents the key conclusions from this research. Page 10

15 Figure 1.1: Adequacy Framework Approach Page 11

16 2 How does the Land Supply Process Work? 2.1 Aims and Objectives The principal aim of this section is to develop an evidence based framework for understanding and analysing the Auckland residential land supply process. This captures the fundamental components of the land supply process including the relevant incentives and drivers of behaviour of land owners and other relevant market participants. The framework will be based on accepted land supply theory, but will reflect the realities of the Auckland Regional market. The key elements of this section are: Description of the regulatory framework that governs the land supply process in the Auckland Region Assessment of how land supply meets demand within the regulatory process Real world rigidities that impact upon the development process. 2.2 Regulatory Processes Governing Land Supply in the Auckland Region Statutory Framework Section 40 (1) of the Local Government (Auckland) Amendment Act 2004 (LGAAA) directs the Regional Policy Statement to give effect in an integrated manner to the Growth Concept in the Auckland Regional Growth Strategy. The Growth Concept seeks that growth is managed by promoting quality, compact urban environments with the intensification of selected high density centres and corridors. To support the management of growth in the region, the ARC formed the Auckland Regional Growth Forum containing representatives of the ARC and the Region s constituent territorial local authorities. The forum functions as a standing committee of the Auckland Regional Council to implement the Regional Growth Strategy. A Memorandum of Understanding (MoU) commits territorial local authorities to the Regional Growth Strategy (vision, outcomes and principles) as well as a sector based planning approach to provide detail on how the Strategy will be implemented at a local level. These Sector Agreements are the key tool for implementing the RGS (identify centres and corridors for intensification). The MUL is intended to help to manage the impacts of development, providing sufficient capacity to cater for growth while protecting the Region s natural resources and significant heritage resources from the adverse effects of development. The RPS provides a framework for extensions to the MUL and to the limits of rural and coastal settlements. Extensions Page 12

17 should not compromise the general direction of containment and intensification or the achievement of intensification within areas that have been identified for growth MUL Extensions in Practice In practice, residential land supply can be expanded by extending the MUL, involving rezoning of land. This can be initiated either by TA s or by TA s in partnership with property owners. Rcent examples of MUL extensions initiated by TAs include Flatbush (in Manukau City) and the more recent inclusion of Hobsonville Airbase, Hobsonville Village and Massey North. As the ARC have the right to veto changes to the MUL under the LGAA 2004, their support is critical. The process for extending the MUL and rezoning land is lengthy, and can typically take in the order of 5 years from initial feasibility investigations to the issue of decisions on the Plan Change applications. This is not out of step with international processes but necessitates the adoption of long planning horizons by local authorities and the development community. There are several generic tasks which are required in the MUL extension process: 1. Landowner Consensus: Council needs to be satisfied that a consensus exists in land owner initiated rezoning requests 2. Feasibility/Technical Investigations: All pertinent issues investigated, with appropriate recommendations. This work establishes the merit of early release and the outcome is a preliminary structure plan 3. Memorandum of Understanding 4. Structure Plan: Used to manage the effects of development plus sequencing of infrastructure 5. Consultation: Occurs through the rezoning process with Council representatives, Iwi, central government, neighbours, legal reviews, inputs into LTCCP s and Annual Plans 6. Plan Changes and Section 32 Reporting: The location of the MUL on maps is changed and District Plans amended to rezone the land. The form and content of District Plan Changes requires careful preparation and discussion with Councils as an iterative process. The time involved can vary widely, depending on whether existing zones are applied, or tailor made zones (and rules) are developed and applied. An example of the length of time it takes to complete various stages is contained in Appendix F. In general, between 4 7 years elapse. The statutory process involves the notification of the Plan Change applications and the call for submissions and further submissions, followed by the hearings and possible appeal Page 13

18 processes. Section 32 of the RMA also requires the TA to prepare a report documenting the reasons, justification, costs, benefits, efficiencies and effectiveness of the proposed MUL extension and District Plan Change. Specialist reports/inputs from the Structure Plan would also be needed as a technical basis to support a specific resource management approach. Table 2.1: Typical Plan Change and Planning Consent Timeframes Year Year 1 Year 2 Year 3 Year 4 Year 5 Quarter Feasibility and Setup 2. Structure Plan 3. Consultation 4. Plan Change/Variation i. Preparation ii.pre notification Processing iii. Notification Periods iv. Officers Report v. Hearing & Decision vi. Appeals 5. Masterplan 6. Resource Consents 7. Construction (subdivision only) While the above discussion relates specifically to extensions to the MUL, it is important to note that essentially the same process is required to establish a new town separated from the urban fringe. More consultation would be involved with nearby municipalities because of obvious overlap and possible diseconomies. 2.3 Land Supply Process within Auckland Region Land Demand and Supply The MUL is an important part of the regulatory framework that governs growth and development in Auckland. In simple terms the MUL aims to curb urban sprawl, its main effects being to constrain the amount of available residential land (especially for conventional household sections). This effect will get stronger as available supply reduces. Page 14

19 It has the potential to lead to undesirable outcomes too, including the opportunity for increasingly scarce land to become tightly held by a relatively small number of owners who can then influence the rate of development and, therefore, the market price for the land. Effectively, land can be drip fed into the market as owners know that demand pressures in Auckland are building all the time, meaning that the price tomorrow will almost inevitably be higher than today. Motu Economic and Public Policy Research found that the release of small parcels of land at a time is supported by Council actions (for instance with regard to infrastructure provision) which promote the development of small quantities of land at a time. This unintentionally strengthens landowners positions and keeps land prices high. Regardless of whether the same amount of land is developed in the long run, the presence of the MUL raises the price of sections developed on the fringe. This in turn has the effect of raising the price of all residential land within the MUL through spatial arbitrage (the process whereby neighbouring sections will tend towards the same price) Land Demand Drivers The demand for residential land in Auckland is driven by a range of factors. Auckland is New Zealand s primary city. It captures a significant portion of the country s growth and the majority of international migrants settle within the region. The key drivers of demand are: Migration international and inter regional Natural growth populations fertility and mortality Demographic characteristics age, sex and ethnicity influence the above but also the type and nature of housing required Preferences for location and housing type Income and wealth The first four aspects define the scale and nature of residential demand, whereas income and wealth define the population s ability to pay for housing choices. In addition to these characteristics longer term trends within the economy, captured within housing cycles define the housing markets overall (supply) response to general economic trends The Housing Cycle The housing cycle or the property cycle describes the manner in which house prices fluctuate in a long run cyclical nature that is generally counter to interest rate movements. In other words, when interest rates are high, households find it difficult to sustain higher mortgages and look to limit their exposure to the interest rates often by selling. This has the effect of depressing residential prices and therefore investment in the residential market. Conversely, when interest rates are low, and there is a relatively low cost Page 15

20 associated with getting into the property market, demand increases and house prices come under pressure to rise. This, in turn, attracts investors into the market to provide new housing stock. In the New Zealand context, it is likely that as monetary policy relaxes it begins a cycle of house price increases and, therefore, increased investment in residential property in general. This is likely to lead to increased pressure on land markets as investors and developers seek new areas and pieces of land to develop and ride the increases. With relatively low interest rates for much of the past 10 years ( at least), there has been significant consumption of vacant residential land within Auckland. Coupled with that has been a sharp rise in house prices. At present, interest rates are high and have been climbing steadily since late 2003 as the Reserve Bank looks to suppress inflation pressures in the market. House prices have been relatively resilient to these increases until quite recently, when average house prices have started to decline. The response in terms of new dwelling permits issued has been more immediate with a steady decline from 8,500 issued in June 2004 to 5,400 issued in June Houses and land are durable products and developers are forward looking, acting today in anticipation of what might happen in the future. This is evidenced in the Auckland market with a number of developers holding large blocks of suitable residential land for long periods of time often decades, before proceeding with actual development. Such land holding is, on one level, speculative and much house building is in anticipation of future demand; so called spec built housing that has been a strong feature of the Auckland housing market. However, in the long run it is seen as a one way bet by market participants who recognise the increasingly scarce nature of land and the reasonably certain projected future growth in demand Demand Growth Migration In the Auckland Region, the majority of growth is attributable to migration (65,000 new migrants between 2001 and 2006 accounting for some two thirds of total population growth), with significant net outflows of existing residents between 2001 and 2006 (18,000, as reported in Auckland s Population, a briefing paper to the Royal Commission on Auckland s Governance by Dr Phil McDermott). However, it is not clear that there is a causal link between the influx of migrants and rising house prices as a recent Motu Research study found that the influx of returning ex pat New Zealanders was more closely linked with increases in house prices than international migrants overall. Page 16

21 The McDermott report highlights a few issues policy makers should heed when assessing future land requirements. Firstly, international migration is forecast to be a significant proportion of Auckland s future population growth up to 44% in Statistics New Zealand s high population forecasts. Secondly, Dr McDermott cautions that New Zealand competes internationally with Australia, Canada and other developed Western Democracies for migrants and that, should New Zealand s competitiveness decline relative to other countries, these flows could drop rapidly. Thirdly, that aging demographics in most western economies mean reduced labour forces and a likely increase in marketing by nations internationally for skilled and semi skilled migrants. International migration is highly volatile and as it plays such a major role in Auckland s future growth, downward shifts in the relative attractiveness of Auckland or New Zealand would translate directly to reductions in the amount of land required to accommodate household growth Demand Growth Natural Increase Natural increase is the other component of dwelling demand growth. Natural population increase is effectively births less deaths over time, therefore is dependent on the age/sex/ethnicity/income characteristics of the local population. As a population ages natural increase declines as mortality rises and fertility drops. In the Auckland Region, there is a strong ethnicity component to natural population growth, with those identifying themselves as NZ European having the lowest fertility rates and the generally younger Maori and Pacific Island communities having the strongest population growth characteristics Rigidities and Blockages Rigidities and blockages in the land supply process can take a number of forms: Urban limits, constrained land supply and leap frogging The presence of the MUL has the potential to greatly enhance TA s ability to control urban sprawl and thus the associated infrastructure, transport and environmental costs. Urban sprawl is subject to negative externalities (and urban agglomeration is subject to positive externalities) and so attempting to constrain sprawl (promote agglomeration) is, in theory, socially optimal. If effective, the MUL would provide sufficient incentives on market participants for intensification. However, if ineffective, it can lead to an acute shortage of land and sharp increases in prices, potentially constraining growth within the region. One possible impact of the MUL is the practice of leap frogging, whereby development occurs some distance away from the metropolitan urban area because of differences in Page 17

22 zoning rules near to, and away from, the urban fringe. In the Auckland region, leapfrogging occurs in a controlled manner resulting in growth around rural towns in Rodney. Significant pressure to develop land immediately north of the MUL is being focused into future urban growth areas around Kumeu Huapai, Riverhead, Orewa and Waimauku as the Countryside living areas are becoming more highly developed. In addition, significant growth at Warkworth and the Rodney coastal settlements of Snells Beach and Algies Bay is strongly linked to growth pressures in the Northern Sector Infrastructure constraints Delays in the funding and development of pipe and roading infrastructure can also constrain growth. This can be reinforced by the sequencing of development as most providers need to develop infrastructure from the core outwards, especially when connections to main trunks are required. Land banking The magnitude of economic rents earned by land owners vary, for example, if the tracts of land owned by the land owner are sufficiently scarce then the land owner has stronger market power. Therefore, land within the urban fringe may purposefully be withheld from early development as a result of optimal inter temporal decision making in a growing urban area. Holding land in a constrained market provides greater market power in the future and ability to set prices. This implies that (in a general sense), an increase in market power among landowners results in a decrease in the pace of development, and a consequent reduction in aggregate land generated benefits. In the Auckland situation, monopolist market power is revealed by the extent of land banking activity within the market (limited evidence of this due to limited assessment timeline, as discussed below), the leapfrogging of development to rural townships and beyond, and the significant residential land price increases in recent years driven by increases on the fringes. Page 18

23 3 Auckland Region Supply 3.1 Auckland Region Capacity for Growth Study 2008 (Theoretical Supply) This study uses the Auckland Regional Council (ARC) Capacity for Growth Study, 2008 as the basis for analysing the adequacy of residential land supply in the Auckland Region. The ARC Study report provides an update of an earlier March 2003 report. The 2008 report is based on raw data current to March ARC has provided DBH with full access to the capacity datasets. The capacity for growth study draws together a range of datasets of residential land supply within the Auckland Region and uses a GIS application to identify, assess and capture sites that are suitable for additional development. This data has been assumed to be accurate. Each parcel of vacant land (or group of parcels in some cases) is identified and captured as an individual row of data and the columns of the database are populated with information on zoning, density, area etc. The key datasets are: 2006 Colour Digital Aerial Photography 2006 Digital Cadastral Database (DCDB) parcel boundaries) Digital District Plan zoning from Auckland Region s Territorial Authorities (TAs) District Plan zoning density assumptions District Plan zoning development controls (yard setbacks, bush protection etc) Dwelling Data (QVNZ) Building Footprints. This information is used to identify the following eight categories of residential land available for development: Vacant Land: any residential zoned land parcel that did not contain any buildings in Vacant Potential Land: any parcel of residential land that has a site size greater than 2,000m² with one or more residential buildings on it and includes a portion of undeveloped or vacant land. Infill (General): any residential zoned site smaller than 2,000m² that can accommodate one or more additional dwellings on the front or rear of the site. Infill Redevelopment: any residential zoned land that is suitable for the removal and replacement of the original house with more than two townhouses (to the maximum permitted density). Page 19

24 Business Land Redevelopment: any business zoned land that makes provision for some residential zoning (i.e. mixed use). The residential component has been calculated as a percentage of the total permitted floor space. Special Areas: capacity in greenfield areas with Structure Plans, such as Flat Bush in Manukau City, Long Bay in North Shore City and Mt Wellington Quarry (Stonefields) in Auckland City. Rural Towns: data on available capacity within the rural towns from the 2008 study has been applied as the best estimate of available land. Rural Residential: covers any rural parcels that are zoned in a manner that allows subdivision to create sites smaller than 8 hectares. The Capacity for Growth Study has identified that there is capacity for approximately an additional 184,450 dwellings (Table 4.1). However, this figure does not include the Infill Redevelopment estimate of 65,051 dwellings. Infill and Infill Redevelopment estimates are not discrete sets, as a single property may be large enough to include a new dwelling on the front or back (pure infill), or as an alternative, the house is able to be removed opening the entire site for potentially more intensive redevelopment. This means that a site can be a candidate for either Infill or Infill Redevelopment. By assuming Infill Redevelopment occurs rather than infill, 2006 capacity increases to 227,220. In reality, a mix of both infill and redevelopment will occur so the true capacity figure will probably lie between these extremes. Table 3.1: Auckland Region Total Household Capacity under current policy 2006 Type of Capacity (ARC) Additional Dwellings: Infill Additional Dwellings: Redevelopment Vacant Land within MUL 59,071 59,071 Infill General 22,280 Infill Redevelopment 65,051 Redevelopment on Business Land 72,367 72,367 Rural towns (estimate) 9,855 9,855 Rural General Rural 1,082 1,082 Rural Countryside Living 19,795 19,795 Total Additional Capacity 184, ,221 Page 20

25 Limitations with Theoretical Supply There are two key limitations with the ARC 2006 measure of theoretical capacity; The ARC measure estimates the stock of capacity and does not take into account the flow of land supply over time, which requires taking account of developer intentions, the stage of the development process that land parcels are at, infrastructure and other timing constraints. It is current to March 2006, and therefore does not account for intervening reductions or additions to capacity to June Real World Constraints The ARC datasets do not include any record of the stage that any identified parcel of land is at in the consent/development process. They do not, for example, distinguish between a vacant parcel of land with subdivision consent in place and a parcel of land with no consents in place. Given the time taken to prepare and obtain consents, this distinction has an obvious bearing on the timing of the subdivided sections (and subsequent homes) coming to the market and available for purchase and occupation. The data does not provide any indication of the timing of major infrastructure investments (water, wastewater, stormwater, etc) that may delay both the availability of portions of vacant land (primarily on the fringes of the MUL) for subdivision and the ability of certain areas of the city to be developed further through infill or redevelopment. For example, infill and redevelopment in certain catchments on the western side of the central isthmus are being held back until the western interceptor wastewater line is completed. The data does not contain information on the length of additional time third parties might add to the consent process. However, the quantum of capacity on larger vacant land parcels has been estimated with care, referencing a range of case studies of recent fringe residential subdivisions. Actual residential densities achieved in these developments have been applied to other potential areas on the urban fringe to provide a relatively accurate measure of residential potential rather than relying on a theoretical maximum as defined by the zoning controls. The ARC datasets provide a record of the quantum of capacity as at March Further information is required in order to project forward the rate of release of this land to market, as the ARC dataset does not record the stage of the land in the consent/development process. Page 21

26 Growth to June 2008 During the period March 2006 and June 2008 a portion of the March 2006 capacity has been consumed (i.e. houses and apartments built). Areas of new capacity have also been added. The March 2006 data, therefore required contemporisation for these changes in supply. 3.2 Summary of Updating Capacity Process, The ARC 2006 database identified some 8,151 parcels of land within the MUL as being vacant, providing estimated capacity for approximately 28,800 dwelling units. In order to update the 2006 ARC data to a 2008 starting point for this study it is necessary to remove residential capacity that has been consumed ( ) and to add capacity that has come on line over the past 2 years. These processes are summarised below and explained in detail in Appendix E Deduction of Capacity / Addition of New Capacity Starting with the ARC s 2006 database, land parcels were removed that have since been developed. This was carried out through a mix of; Site Visits to visually check large capacity areas for completed units In house information in respect of developments that are known to be completed; Council consent tracking databases to identify sites for which Code Compliance Certificate units 1 had been issued. The first two processes covered some 881 parcels or 70% of available vacant land within the MUL, while the Council consent tracking data covered an additional 1,100 parcels and a further 12% of vacant land. Areas not able to be covered by these processes (due to data constraints) have been carried through from the 2006 dataset with some allowance for interim consumption and development at realistic densities. These are conservative assumptions in terms of consumption and are likely to result in a small overstatement of capacity. However, the difference is likely to be less than a few percentage points as this group of properties represents less than 14% of total residential land of which only a relatively small portion would have been consumed Quantity of Residential Capacity consumed 1 The term unit is utilised in this report to refer to a finalised product which can be occupied, that is a house, flat, apartment, townhouse, etc. Page 22

27 Between 2006 and 2008 the amount of capacity on vacant and vacant potential land dropped from 28,798, as reported in the Capacity for Growth Study 2008, Interim Report (ARC) to 25,273. This represents a decline of some 3,500 over two years but contains within it differences in assessment of capacity making comparisons problematic. Capacity consumption over the past two years for other forms of capacity are also difficult to determine. Business areas and rural capacity from 2006 has been adopted as the 2008 figure (with some allowance for error and superior data). Infill has been carefully calculated by the ARC for 2006 using aerial photos and GIS systems, but data on its consumption (primarily through building consent information) is impossible to match to individual properties with accuracy. In addition, the range of capacity outcomes from the different development options (infill and redevelopment) is large, making consumption over the past two years irrelevant Quantity of Residential Capacity added ( ) Waitakere City The ARC 2006 datasets were completed prior to the rezoning (as part of the Local Government (Auckland) Amendment Act (2004) process) of three areas of land in Waitakere City, which together bring significant additional vacant residential capacity to the Region. The three areas are as follows: Hobsonville Airbase (Plan Change 13, 3,110 units); Hobsonville Village (Plan Change 14, 110 units); Massey North (Plan Change 15, 1,330 units). In addition to these three areas, the rezoning of New Lynn (Plan Change 17) added capacity through redevelopment opportunities for intensive housing (to a total of 6,000 units). Capacity for each of these four areas was added to the dataset. The Hobsonville Airbase capacity (and staging) was based on an interview with Hobsonville Land Company (HLC) and knowledge of the development process due to previous involvement in this development. Capacity (and staging) for Hobsonville Village and Massey North was based on Waitakere City Council (WCC) projections 2. Revised New Lynn projections were based on discussions with Waitakere City Council Officers Franklin District 2 Projections included in Draft Transport Audit for the Auckland Regional Transport Authority of Waitakere City Council s proposal to develop the Massey North to Hobsonville Corridor dated March 2006, prepared by WCC. Page 23

28 The ARC dataset is limited to the Metropolitan Urban Limits (MUL) and does not provide any data for Franklin District, as this falls outside of MUL. The brief, however, requires that the part of the District within the Auckland Region be included,. A meeting was held with Franklin District Council (FDC) staff who advised that the only areas of current residential capacity are located within Pukekohe and Waiuku. Further information on capacity in those two areas was obtained by site visits and supplemented through discussions with Council Officers to complete the 2008 assessment of regional capacity Infill Sites in All TAs Infill has been treated as though the capacity exists today and will be consumed in much the same way as small parcels of vacant land are. This means that as population comes on line over time it is allocated to vacant capacity and infill capacity in much the same manner. Total supply has been constrained region wide as it is not the case that every property available for infill will be developed. The model framework allows for a number of scenarios to be run. As a base case, an assumption that some 20% of available capacity will not come to market, has been made. In addition, a portion (20%) of infill capacity has been set to redevelopment infill rather than general infill. Redevelopment provides a higher yield than infill but is not as easy as it often requires aggregation of individual lots to create sufficient footprint to allow redevelopment. In total the Infill/Redevelopment category contributes around 23,400 dwellings to total regional capacity. More detail on how infill is treated in contained in Appendix E that accompanies this report Capacity in Business Centres Capacity in business centres has been adopted (in the main) from the ARC study due to difficulties in quantifying theoretical residential capacity maximums and therefore how much capacity has been consumed. Household growth and population change over time drives the actual uptake into the future In total business areas are anticipated to deliver approximately 75,260 dwelling units (assumed to all be high density apartments). This is a significant 41% of total regional capacity and represents a significant risk for two reasons; Page 24

29 Firstly, space in centres is the most highly competed for by a range of uses (office, retail, services). This means the theoretical 30% of built floorspace devoted to residential may not eventuate as residential may fail to compete with other uses for the space, especially if housing preferences do not shift sufficiently to ensure a high level of demand and therefore reasonable returns for developers. Secondly, having such a large share of total capacity in business areas is likely to require significant dwelling and location preference change to ensure it meets the future housing needs of the population. Currently these areas account for 12% of total households in the region Rural Capacity For the purposes of this report, rural capacity is split between its three components; rural townships (with future urban zone areas), rural countryside living (capacity on the MUL edge that is able to be subdivided into lots of 1ha, this includes farm park type developments) and general rural (effectively farm locations able to be occupied and utilised as farms) within each TA. Rural capacity was originally to be excluded from the scope of work for this project as the focus was on the ability of residential land within the MUL to accommodate Auckland s growth to support the Compact City outcome. However, it is important to account for rural capacity as it represents a significant percentage of theoretical capacity. This raises two key questions; firstly, whether theoretical rural capacity can be converted into actual capacity and secondly, whether it should be given Auckland s focus on a Compact City form. In total rural areas contribute some 30,570 dwellings to regional capacity. This is made up of; Rural Townships 9,690 Countryside Living 19,800 Rural General 1, Residential Capacity, 2008 Through the combination of site inspections, discussions with Council Officers, and (latterly) reference to the TA s tracking databases and modelling (as described above) an up to date (1 June 2008) record of capacity was derived for all of the major areas of capacity identified in the ARC datasets. Page 25

30 Local Authority Table 3.2 Auckland Residential Capacity 2008 Vacant and Vacant Potential Infill/ Redevelopmen Structure Plan t Business Areas Rural Total Capacity Rodney District 2, ,447 2,758 19,223 32,464 North Shore City 7,429 3,805 2,900 10, ,438 Waitakere City 5,162 4,852 5,992 9,051 1,571 26,628 Auckland City 4,125 6,410 2,900 44,476 2,629 60,540 Manukau City 4,994 6,719 12,060 6,143 2,484 32,400 Papakura District 949 1,181 2, ,129 Franklin District ,379 3,712 5,091 Auckland Region Total 25,273 23,388 34,199 75,259 30, ,689 *Infill total in this table does include a portion (20%) that might be redeveloped at a higher level of intensity In total the region has residential capacity for 188,690 dwellings under the current zoning regime. The majority of this capacity resides within Auckland City (60,540 almost 1/3 rd of the regional total). Rodney District has the next highest capacity at 32,460 however the majority of this is rural (19,200). In terms of locational type, the majority of capacity resides within Businesses areas (75,300 or 41%). Rural capacity accounts for a significant 17% of total capacity. Note that the figures in Table 3.1 account for all future capacity rather than that actually available in Table 3.1 overstates the current capacity, for example, the majority of Structure Plan capacity is to come on line over the next 5 10 years and is not ready to be occupied in 2008, whereas the majority of the infill/redevelopment capacity is envisaged under the current zoning regulations and would be able to be developed immediately. Note also, that this estimate includes the totals for the structure plan areas and all vacant potential regardless of when it is expected to come to market. The base run assumes that a portion of infill capacity is redeveloped rather than being used for infill. The base run assumes that General Infill will be adopted by 80% and redevelopment by 20%. This is a realistic figure set as a general guide, however it will vary by TA and over time. Sensitivity testing has been carried out to assess the implications of altering this assumption (see Scenario 1 below). To attempt to determine the proportion of redevelopment infill, patterns of redevelopment have been studied covering the past 10 years (as reflected in the building consents). Matches between rates of redevelopment and land to capital value ratios were explored, however the data did not strongly support the contention that properties with either a high land to capital ratio or some other defining characteristic, would be redeveloped over Page 26

31 others. Without a sound basis for distinguishing between areas or properties, it was not possible to be definitive, therefore a more general approach was adopted Capacity by Density Type The majority of available capacity is classified as high density, that is it is associated with less than 200sqm of land. In total 85,136 units are classified as high density, 47,000 in Auckland City alone (57%), the majority of these in business areas (Table 3.3). Table 3.3 Auckland Residential Capacity by Density Type 2008 Local Authority Conventional Medium High TOTAL Rodney District 23,658 5,136 3,670 32,464 North Shore City 4,952 7,083 13,402 25,438 Waitakere City 12,104 4,125 9,834 26,062 Auckland City 8,387 4,676 47,003 60,067 Manukau City 9,964 14,189 9,288 33,441 Papakura District 4,529 1, ,127 Franklin District 3,712-1,379 5,091 Auckland Region Total 67,306 36,247 85, ,689 Conventional density accounts for 36% of the total (67,300 households) of which 35% reside within Rodney District (under the full capacity scenario). This again represents significant risk as it involves a significant portion of countryside living, directly contrary to the aims of a Compact City Capacity Sensitivity Testing In order to test the sensitivity of the capacity outcomes to the above assumptions a number of alternative scenarios have been run. These aim to test the effect of: A higher share of redevelopment as opposed to general infill. It is anticipated that redevelopment is most likely in the major metropolitan cities first. In Auckland, North Shore, Waitakere and Manukau cities, redevelopment has been set to 50% of total potential compared with 20% in the base case. Rodney, Franklin and Papakura have been left as is. Rural capacity not being a strong option to cater for growth. This is made up of two parts, firstly general rural capacity is withheld and countryside living is included at 50% of theoretical capacity. This brings it more in line with recent consumption trends (1996, 2001, 2006) and makes the assumption that this will continue into the future rather than having the full amount available to accommodate households seeking conventional density. Page 27

32 A third scenario simply looks at the capacity within the MUL plus rural towns. This is an extreme stance and represents the minimum. Page 28

33 Scenario 1: High Redevelopment Scenario Under the high redevelopment scenario an additional 10,000 dwellings are anticipated within the MUL. Total capacity rises to 198,800 with medium density now accounting for 22% of total capacity (up from 19%). In total, this shift adds around 5% to total regional capacity, which is not a particularly significant shift given the significant difference in behaviour it would entail on the part of developers (half of all infill occurring as redevelopment). Table 3.4: High Redevelopment Scenario Residential Capacity 2008 Local Authority Vacant and Vacant Potential Infill/ Business Structure Plan Redevelopment Areas Rural Total Capacity Rodney District 2, ,447 2,758 19,223 32,464 North Shore City 7,429 6,023 2,900 10, ,656 Waitakere City 5,162 7,018 5,992 9,051 1,571 28,794 Auckland City 4,125 9,517 2,900 44,476 2,629 63,647 Manukau City 4,994 9,296 12,060 6,143 2,484 34,978 Papakura District 949 1,181 2, ,129 Franklin District ,379 3,712 5,091 Auckland Region Total 25,273 33,458 34,199 75,259 30, ,759 Table 3.5: High Redevelopment Scenario Capacity by Density Type 2008 Local Authority Conventional Medium High TOTAL Rodney District 23,658 5,136 3,670 32,464 North Shore City 5,684 8,559 13,413 27,656 Waitakere City 12,768 5,601 9,860 28,228 Auckland City 9,158 6,898 47,119 63,174 Manukau City 11,038 15,675 9,306 36,019 Papakura District 4,529 1, ,127 Franklin District 3,712-1,379 5,091 Auckland Region Total 70,545 42,906 85, ,759 Page 29

34 Scenario 2: Restricted Rural Capacity Scenario The effect of restricting rural capacity is significant, especially for the supply of conventional density land. The base run provides capacity of 188,690 dwellings. By restricting rural capacity to 50% of Countryside living and excluding general rural, some 11,000 dwellings are removed from Base Case capacity all of which are conventional density (6% of the base total). This could arise should councils with rural capacity seek to limit rural sprawl in an effort to develop rural townships and promote intensification within the MUL. Significant changes occur in Rodney District as capacity there is heavily rural (from 32,500 to 25,400 or down by 22%). Local Authority Table 3.6: Restricted Rural Scenario Residential Capacity 2008 Vacant and Vacant Potential Infill/ Business Structure Plan Redevelopment Areas Rural Total Capacity Rodney District 2, ,447 2,758 12,190 25,431 North Shore City 7,429 3,805 2,900 10, ,233 Waitakere City 5,162 4,852 5,992 9, ,842 Auckland City 4,125 6,410 2,900 44,476 2,072 59,983 Manukau City 4,994 6,719 12,060 6,143 1,643 31,559 Papakura District 949 1,181 2, ,773 Franklin District ,379 2,510 3,889 Auckland Region Total 25,273 23,388 34,199 75,259 19, ,709 Table 3.7: Restricted Rural Scenario Capacity by Density Type 2008 Local Authority Conventional Medium High TOTAL Rodney District 16,625 5,136 3,670 25,431 North Shore City 4,747 7,083 13,402 25,233 Waitakere City 11,319 4,125 9,834 25,277 Auckland City 7,830 4,676 47,003 59,510 Manukau City 9,123 14,189 9,288 32,600 Papakura District 4,173 1, ,771 Franklin District 2,510-1,379 3,889 Auckland Region Total 56,327 36,247 85, ,709 Scenario 3: No Rural Capacity Scenario Under the no rural capacity scenario, only capacity within either the MUL or the rural towns (plus their Future Urban Areas) is available for residential growth. This scenario sees total capacity drop to around 167,800. This scenario sees some 87% of capacity residing within the MUL and conventional density dwellings only making up 28% of total capacity. Page 30

35 Local Authority Table 3.8: No Rural Scenario Residential Capacity 2008 Vacant and Vacant Potential Infill/ Business Structure Plan Redevelopment Areas Rural Total Capacity Rodney District 2, ,447 2,758 5,157 18,398 North Shore City 7,429 3,805 2,900 10,894-25,028 Waitakere City 5,162 4,852 5,992 9,051-25,057 Auckland City 4,125 6,410 2,900 44,476 1,864 59,775 Manukau City 4,994 6,719 12,060 6,143 1,181 31,097 Papakura District 949 1,181 2, ,588 Franklin District ,379 1,491 2,870 Auckland Region Total 25,273 23,388 34,199 75,259 9, ,812 Table 3.9: No Rural Scenario Capacity by Density Type 2008 Local Authority Conventional Medium High TOTAL Rodney District 9,592 5,136 3,670 18,398 North Shore City 4,542 7,083 13,402 25,028 Waitakere City 10,533 4,125 9,834 24,491 Auckland City 7,622 4,676 47,003 59,302 Manukau City 8,661 14,189 9,288 32,138 Papakura District 3,988 1, ,586 Franklin District 1,491-1,379 2,870 Auckland Region Total 46,429 36,247 85, , Analysis of Supply Total currently available capacity has been arrived at by undertaking the above assessment and for land where no assessment has been able to be carried out, the 2006 capacity figure has been relied upon factored down slightly in the case of vacant land, to take into account the density land plots (in the same zones) have been developed to. For example, if a zone has a minimum density of 350sqm per lot, yet development in recent years in these areas has resulted in 450sqm averages, the yield is not as great. These differences are expressed in estimates where actual numbers are not known. This process provides a more realistic assessment of capacity than simply relying on the theoretical capacity captured in the 2006 data set. Details on how this has been carried out are contained in the Appendix E. As the data is collected and analysed at a property and area unit level, it is possible to assess the within MUL versus outside MUL patterns of capacity. Table 3.10 splits capacity from the Base Run between density categories of dwellings within and outside the MUL. It highlights the high proportion of conventional housing capacity located outside the MUL (50%). As development of this would be directly contrary to the aims of the Regional Growth Strategy and the Compact City model, this represents a significant challenge to the regions councils Page 31

36 to adequately accommodate family household growth (those most disposed to conventional density dwelling arrangements) over the next 30 years. In addition, it is worth noting that urban and rural capacity differs significantly in character, meaning, it cannot be treated in the same manner. It is not the case that a family looking to locate in a standalone house in a suburban location would switch to a rural location. Given research published by the ARC 3 households are more willing to seek higher density environments rather than move, especially to significantly different areas (urban to rural). This is especially true for families looking to stay within particular school zones or the elderly looking to remain in the same suburb. Also, there is likely to be resistance from established rural communities to the growth envisaged in rural areas under the Base Run scenario, even in cases where zoning permits more intensive forms of development due to the alterations in rural character it would bring. Infrastructure constraints are also likely to hinder rural and rural township development (the delays associated with wastewater constraints in Kumeu/Huapai in Rodney are a prime example of this). Zoning of rural residential is at odds with both the Compact City model and councils aims of intensifying in and around rural townships. One local authority with large rural capacity indicated they were considering tightening the rules on rural development to reduce the impact of low intensity life style development. The effects of this type of approach on regional capacity overall have been modelled in Scenario 2 and 3 above. In total, 77% of capacity lies within the MUL and 23% outside. This is more in line with the stated aim of the ARC to accommodate some 75% of future growth within the MUL. However, as stated elsewhere in this report, in order to achieve that significant preference changes are required to encourage a wider range of household types to adopt higher density living arrangements. 3 Building a Better Future, (2000) Research Solutions as summarised in David Mead s Regional Intensification Intensive housing demand and supply issues (2007) Page 32

37 Table 3.10 Total Residential Capacity by Density Category and TA, 2008 Within MUL Outside MUL Local Authority Conventional Medium High Conventional Medium High TOTAL Rodney District 3,913 4,590 1,563 19, ,108 32,464 North Shore City 4,542 7,083 13, ,438 Waitakere City 9,822 3,962 8,438 2, ,395 26,062 Auckland City 6,006 4,662 46,091 2, ,067 Manukau City 7,100 12,162 8,095 2,864 2,027 1,193 33,441 Papakura District 2, , ,127 Franklin District ,712-1,379 5,091 Auckland Region Total 34,265 33,286 78,044 33,041 2,961 7, ,689 Share by Density Class 50.9% 91.8% 91.7% 49.1% 8.2% 8.3% Total Share 77.2% 22.8% 100.0% The vast majority of medium and high density capacity resides within the MUL (92% for each). 3.5 Identification of Current Stage in the Consent/Development Process Having established the quantum of residential land supply it is necessary to understand how close to market the supply is, before questions of adequacy can be properly answered. Auckland s residential land is in various states of readiness to be developed, so each parcel has been coded according to the stages in the process identified in Table Once average timings for progress through each stage are added, a consistent robust way of assessing when land will come on line can be developed. More detail on the coding process is contained in Appendix E. Table 3.11 Current Stage in Consent/Development Process (Milestones) and Periods Between Stages Current Stage in Consent / Development Process (Milestones) and Periods Between Stages Code Stage in Consent / Development Process Code Periods Between Stages (Milestone) 1 Original 2006 Land Parcel 1 The preparation of Comprehensive Development Plan (CDP) (or similar) documentation. 2 Comprehensive development plan lodged 2 The CDP (or similar) processing period. 3 Comprehensive development plan approved Page 33

38 Current Stage in Consent / Development Process (Milestones) and Periods Between Stages Code Stage in Consent / Development Process Code Periods Between Stages (Milestone) 3 Design / documentation for subdivision consent. 4 Subdivision consent lodged 4 Subdivision consent processing period. 5 Subdivision consent approved 5 Subdivision construction period / associated engineering approvals / tendering up to release Section 224(c) Certificate C issued 6 Period between release of Section 224(c) Certificate up to issue title. 7 Titles issued 7 Design/documentation for land use consent for buildings or combined subdivision/land use consent. 8 Land use / combined consent lodged 8 Land use or combined subdivision/land use consent processing to approval. 9 Land use / combined consent approved 9 Building design period for 1 st Building Consent (BC) to enable construction to commence. 10 Building Consent lodged 10 Building consent processing period to enable construction to commence. 11 Building Consent approved 11 Building construction period to first inspection (foundations). 12 First building inspection (foundations) 12 Building construction from first foundation inspection to pre lining inspection (or proxy). Page 34

39 Current Stage in Consent / Development Process (Milestones) and Periods Between Stages Code Stage in Consent / Development Process Code Periods Between Stages (Milestone) 13 Pre lining inspection 13 Building construction from pre lining inspection (or proxy) to CCC (or proxy). 14 Code Compliance Certificate issued and where applicable a subsequent subdivision consent lodged. 14 Unit title/cross lease/fee simple subdivision processing period (if further subdivision required) to release of s224(c). 15 Subsequent subdivision consent approved 15 Average time to gain survey plan (LINZ) approval and issue of titles for further subdivision. 16 Subsequent titles issued. 224(c) issued Vacant Capacity by Development Stage All vacant residential land (either vacant or vacant potential, including that contained within Structure plan areas) has been coded against this framework and is presented in Table 3.11, below. This table indicates that approximately 30% of the land covered has entered the consent system (covering approximately 17,460 units) with the largest portion at the subsequent subdivision consent lodged stage. Note this is made up primarily of parcels in Structure plan areas. The largest portion of vacant capacity sits at the subdivision consent approved stage. Note that other forms of capacity (infill, business land, rural) are not subject to the same processes and do not fit within this framework. Page 35

40 Table 3.12: Auckland Vacant by Consent Process Stage, 2008 Stage in Consent Process Vacant Capacity Structure Plan Areas TOTAL Original 2006 Land Parcel 20,475 20,894 41,369 Comprehensive development plan lodged Subdivision consent lodged ,528 Subdivision consent approved 1, , C issued Titles issued 426 1,879 2,305 Land use / combined consent lodged 381 2,069 2,450 Land use / combined consent approved 904 1,068 1,972 Building consent lodged Building consent issued First building inspection Code Compliance Certificate Subsequent subdivision consent lodged 280 4,655 4,935 Subsequent titles issued #N/A TOTAL 25,272 33,554 58,825 Having established where each parcel of vacant land currently is in the consent/development process, and having exposed it to the dimensions of influence, our model is able to express, in months from 1 June 2008, the overall timing of release of the parcels of land based on the preceding factors. 3.6 Model Months to Market It is necessary to forecast the time that it will take each parcel of land to reach the market as completed liveable units. In order to do this, each parcel was exposed to the dimensions of influence which determine the duration of this process. These dimensions of influence are: The consent/development process Landowner aspirations Infrastructure constraints Public and private agencies Property cycle Construction season. In practice it takes a shorter amount of time for a small two lot subdivision to work through the consenting system than a large 50 lot subdivision, simply because the effects of the Page 36

41 development on the environment, on policy goals local authorities might have, on neighbours and infrastructure providers, are likely to be far less. Details of how land parcels are coded are contained in Appendix E, the process is summarised below. Consent / Development Process Each parcel of land has been coded according to its current stage within the consent/development process. All other things being equal, the components and total period that this process takes differs depending on the type of development involved. In order to model this consent/development process the following information has been incorporated into the model: 1. The current stage that the parcel is in the consent/development process; 2. The remaining steps of the consent/development process that the land will have to pass through to completion of a known or assumed development. This was developed in consultation with Council officers and included details of the time periods associated with each step (detailed in Appendix E); 3. The type of development proposed or likely. A generic list of 24 types allowing compatibility between TA s; 4. The expected durations of, and interactions between, the remaining time periods. This allowed distinctions to be made between notified and non notified and critical paths to be established. Distinctions between the development process in areas The remaining time periods (i.e. other than Council processing timeframes) identified in the tables relate (in summary) to: Design/documentation periods Construction periods Certification and LINZ periods. Landowner Aspirations The Development Framework recognises that developer aspirations may override the timings to release that could otherwise be assumed if one relied only on the sequential application of design/consent processing/construction timeframes. A landowner could obtain consents for a project but then (for a number of reasons) choose to hold the Page 37

42 land in an undeveloped state rather than proceed to develop it in accordance with the consent. Similarly, if there are no consents in place for a landholding, then it is not possible to confidently estimate the timing of development and ultimate release to market of units on that land, as the date of commencement of the process is not known. Interviews were undertaken with landowners and developers of some of the largest vacant landholdings within the region. Area not covered, an assumed aspiration has been applied based on the timings of areas in close proximity. Infrastructure Constraints In greenfield areas, infrastructure is typically extended and developed at the same time as the development of the land. In these situations, infrastructure is put in place as blocks of land are developed. There are situations where blocks cannot be developed until such time as piped infrastructure is provided to the boundary, and this is reliant on the development of the downstream blocks. These issues are specific to each individual block of land and dependant on timing of surrounding development. They are progressively resolved as development proceeds, and this is generally market driven. Irrespective of landowner aspirations and permissive zoning provisions, in other areas where there are infrastructure constraints, development opportunities are limited until constraint is resolved. Examples may include; wastewater infrastructure at capacity. Further development (anticipated by the zoning) is precluded until additional capacity is provided. trunk piped infrastructure that has not yet reached the boundary of a greenfield area otherwise available for development, and the extension of that pipe is beyond the control of the landowners. Information provided by TA Assets staff as to actual constraints, and the expected timing of resolution informed these figures. Public and Private Agencies Public bodies such as Central or Regional Government agencies (Transit New Zealand, the ARC, etc) get involved in the consent process as submitters, and can delay development projects through Appeals. This tends to be more typical in the Auckland context for higher density developments in established neighbourhoods. Page 38

43 Property Cycle It is worth noting that at the time of preparing this report, there are a number of subdivisions which are consented but did not proceed to construction in the 2007/2008 construction season due to the current market conditions and funding issues (a result of the credit crunch ). Based on discussions with developers a delay of between 6 or 12 months from June 2008 would be appropriate and has been incorporated into the figures. Construction Season It is not possible to develop residential land all year round in the Auckland environment. The heavy rains of winter combined with reasonably clay based soils makes operating heavy earth moving equipment impossible. The Auckland earthworks season typically begins on the 1 st of October and runs through till the next winter. We have based the timeframes of the rezoning process on our own experience (acting as principal consultants to landowner/council partnerships) of rezoning in substantial areas of land in Hingaia, Takanini (Papakura District) and currently Pokeno (Franklin District). These typical timeframes are set out in Table 3. The overall process takes between 3 to 5 years in our experience. 3.7 Conclusions Given the limitations described above, it is still clear that Vacant, vacant potential and infill capacity has reduced over the past two years. In total the 2006 study indicated capacity for 49,110 dwellings in these areas, whereas our 2008 is approximately 45,550 a drop of 3,560. Even though capacity in the Structure Plan areas increased by approximately 3,900 dwellings, total capacity dropped by 3,410 between the two studies (once other sources are accounted for). Page 39

44 4 Auckland Region Demand, Housing Demand Model This study uses ME s proprietary Housing Demand Model to estimate future demand for residential dwellings in Auckland Region to The drivers of housing demand are twofold. Firstly, population growth is stimulated by immigration both at international and domestic levels, as well as growth within the existing population. The second driver of demand is the change in the housing demand characteristics of households as they move through a life cycle. As households age, become wealthier, or increase in size, the preference for a specific dwelling type changes. In addition, as the city evolves and matures, under a Compact City growth future households will adopt more intensive living arrangements than have been the norm in the past. Auckland will become more like international cities over time. The model captures these trends by utilising both historic dwelling preferences for different types of households by age, size and income and by adopting a scenario that trends household preferences towards those found in the larger Australian cities (Melbourne and Sydney). At the Auckland Region level, households are aging, becoming wealthier and decreasing in size, and these trends lead to an overall change in preferences for dwelling types. 4.2 Housing Demand Model Methodology and Datasets The output of the model is TA level projections of demand for residential dwellings by density (i.e. conventional, medium and high) for 47 household types (defined by household composition, household income and age of head of household) from 2006 (base) to 2031 at five year intervals. The demand projections are driven by Statistics New Zealand s (SNZ) TA level population projections. 4 SNZ produces three projection series (i.e. low, medium and high) using different combinations of fertility, mortality and migration assumptions for each area, and suggests that the medium projection series is the most suitable for assessing future population growth. The model can be run for all three projection series, however the output presented in the body of this report is based on the medium projections series. Output based on the high projection series is presented as Appendix B, as part of a broader sensitivity analysis. 4 Subnational Population Projections by Age and Sex, 2006(base) 2031 Page 40

45 Dwelling preferences (by density for 47 household types) are inferred from an analysis of historical census data for Auckland Region and, in the case of the Melbourne Sydney scenario, are supplemented with an analysis of revealed dwelling preferences for households in Melbourne and Sydney. 5 Combining the resulting household and dwelling preference projections allows for detailed projections of demand for dwellings that capture the changing dynamics associated with household life cycles and the evolution of Auckland Region under a Compact City growth future. 4.3 Demand for Dwellings in Auckland Region, 2008 to 2031 There are currently approximately 446,000 dwellings in the Auckland Region (Table 4.1). Around 86% of total dwellings are located within the four main urban TAs (North Shore, Waitakere, Auckland, Manukau), with the strongest concentration (33%) in Auckland City. Table 4.1 Estimated Dwellings in Auckland Region by Territorial Authority, 2008 Territorial Authority Dwellings % Rodney District 34, North Shore City 73, Waitakere City 63, Auckland City 146, Manukau City 97, Papakura District 15, Franklin District 15, Auckland Region 445, Based on the medium population projection, it is anticipated that there will be demand for an additional 210,610 dwellings in the region between 2008 and 2031 (Table 4.2). Over this time, Auckland City (31%) and Manukau City (27%) are expected to share over half of the growth in demand for dwellings in Auckland Region (Table 4.3). 5 Although census data does not allow a direct comparison of dwelling preferences in Auckland, Melbourne and Sydney, a general pattern can be observed. In 2006, separate houses comprised 76% of the residential dwelling stock in Auckland Region compared with 73% in Melbourne Statistical Division and 64% in Sydney Statistical Division. The Melbourne Sydney scenario projects changes in dwelling preferences using the rates of change observed between the 2001 and 2006 censuses and is based on Auckland Region rates in the short run, Melbourne rates in the medium run and Sydney rates in the long run. Page 41

46 It is important to note that these projections do not take into account the detailed assessment of available capacity undertaken in Section 4. They have been treated as an input into the adequacy of land supply modelling. Table 4.2 Incremental Demand for Dwellings in Auckland Region by Territorial Authority, 2008 to 2031 (based on Medium Population Projection) Territorial Authority Rodney District 2,067 3,689 3,960 3,989 3,739 17,446 North Shore City 3,590 6,495 6,552 6,561 6,331 29,529 Waitakere City 3,542 6,183 6,387 6,424 6,320 28,856 Auckland City 7,981 14,371 14,331 14,262 13,809 64,753 Manukau City 6,563 11,970 12,503 12,681 12,815 56,532 Papakura District 697 1,250 1,305 1,292 1,313 5,856 Franklin District 929 1,630 1,739 1,717 1,624 7,638 Auckland Region 25,369 45,587 46,777 46,926 45, ,610 Table 4.3 Percentage Share of Growth in Demand for Dwellings in Auckland Region by Territorial Authority, 2008 to 2031 (based on Medium Population Projection) Territorial Authority Rodney District North Shore City Waitakere City Auckland City Manukau City Papakura District Franklin District Auckland Region For the purpose of this exercise, dwellings have been broadly classified as either conventional (separate house land density of 450sqm or more), medium density (two or more flats/units/townhouses/apartments/houses joined together in a single storey building with land density of between 200sqm/unit and 450sqm/unit) or high density (two or more flats/units/townhouses/apartments/houses joined together in a multi storey building with land density of less than 200sqm/unit). Typically, as the supply of land for residential development becomes constrained, higher density dwellings receive increasing general acceptance. The rate at which this intensification occurs varies, therefore two scenarios have been developed. The first scenario is merely a projection of the historical trend evidenced in Auckland Region over the past five years. The second (and preferred) scenario is based on the evidence of Melbourne and Sydney, which are considered suitable examples of cities that are further down the track of higher density residential development. Page 42

47 Under both scenarios, conventional housing is expected to be the most sought after dwelling type in the region. Under the historical scenario, conventional dwellings will comprise 56% of growth in demand for dwellings, while medium density dwellings will comprise 24% and high density dwellings will comprise 20%. Under the Melbourne Sydney scenario, conventional dwellings will comprise 48% of growth in demand for dwellings, while medium density dwellings will comprise 29% and high density dwellings will comprise 22%. Detailed tables showing growth in demand for dwellings by dwelling type based on the low, medium and high population projections are shown in Appendix A. Table 4.4 Incremental Demand for Dwellings in Auckland Region by Dwelling Type, 2008 to 2031 (based on Historical Intensification) Dwelling Type Conventional 15,230 27,201 26,061 24,881 25, ,572 Medium Density 5,044 9,786 11,177 12,150 11,807 49,964 High Density 5,094 8,600 9,539 9,895 8,946 42,074 Total 25,369 45,587 46,777 46,926 45, ,610 Table 4.5 Incremental Demand for Dwellings in Auckland Region by Dwelling Type, 2008 to 2031 (based on Melbourne Sydney Intensification) Dwelling Type Conventional 14,147 23,883 22,196 20,822 20, ,022 Medium Density 6,053 12,179 13,918 15,076 14,490 61,716 High Density 5,169 9,525 10,663 11,029 10,486 46,872 Total 25,369 45,587 46,777 46,926 45, ,610 Page 43

48 5 Adequacy of Supply 5.1 Reconcile Supply & Demand Figures by Time Period In order to address the question of Adequacy of Supply in Auckland Region, it is necessary to match the supply of residential land (translated into dwelling units) with the anticipated growth in demand (translated into household growth). This will allow identification of choke points, both in terms of timing of release or availability of land, and location. In broad terms, there are reasonably significant constraints in the market that manifest before In total, the region is expected to increase household numbers by 210,600 to 2031 under a medium growth future and some 254,000 under the high growth scenario, whilst total capacity (as modelled under the Base Run assumptions) is only 188,690 (Table 5.1), leaving a shortfall of between 21,900 to 65,300 at the regional level. However, the constraints are felt much sooner in some TAs (North Shore City, Auckland City and Manukau City) and in particular in the conventional density class. Table 5.1: Auckland Region Residential Capacity (Units) by TA Local Authority Conventional Medium High TOTAL Rodney District 23,658 5,136 3,670 32,464 North Shore City 4,952 7,083 13,402 25,438 Waitakere City 12,104 4,125 9,834 26,062 Auckland City 8,387 4,676 47,003 60,067 Manukau City 9,964 14,189 9,288 33,441 Papakura District 4,529 1, ,127 Franklin District 3,712-1,379 5,091 Auckland Region Total 67,306 36,247 85, ,689 Table 5.2: Auckland Region Medium Residential Household Growth by TA, Territorial Authority Total Rodney District 2,067 3,689 3,960 3,989 3,739 17,446 North Shore City 3,590 6,495 6,552 6,561 6,331 29,529 Waitakere City 3,542 6,183 6,387 6,424 6,320 28,856 Auckland City 7,981 14,371 14,331 14,262 13,809 64,753 Manukau City 6,563 11,970 12,503 12,681 12,815 56,532 Papakura District 697 1,250 1,305 1,292 1,313 5,856 Franklin District 929 1,630 1,739 1,717 1,624 7,638 Auckland Region 25,369 45,587 46,777 46,926 45, ,610 Page 44

49 In total, the region is anticipated to reach 656,350 households by 2031 under a medium growth future (the most likely outcome) with Auckland City topping 200,000 households some time in 2027 and Manukau City having more households than Auckland City currently does by 2031 (Table 5.3) Table 5.3: Auckland Region Total Households (Medium Growth) by TA, Territorial Authority Rodney District 34,033 36,101 39,790 43,750 47,740 51,479 North Shore City 73,815 77,405 83,900 90,452 97, ,344 Waitakere City 63,621 67,164 73,346 79,733 86,157 92,477 Auckland City 146, , , , , ,394 Manukau City 97, , , , , ,877 Papakura District 15,135 15,831 17,081 18,386 19,678 20,991 Franklin District 15,148 16,077 17,707 19,445 21,163 22,787 Auckland Region 445, , , , , , Allocation Approach Household growth has been allocated to local areas in order to understand the unevenness of under supply across the region and to highlight where the pressure points might occur. The allocation of demand for dwellings (by dwelling type) to areas with capacity for development takes place in two parts at TA level then at local area unit level. That allocation of households to dwellings algorithm makes an assumption that households will seek to preserve their dwelling type preferences rather than their location type preferences, at the point where they cannot both be met locally. In addition to the intensification incorporated in the demand projections, the model has been tested for its sensitivity to a popular acceptance of intensification when demand for conventional housing in a desired location exceeds supply. The ARC Residential Research and Monitoring Programme s report on Building a Better Future: Intensification Review (Research Solutions, 2000) gives some insight into people s attitudes to intensive housing. When survey respondents (residents of intensive housing, as well as residents of surrounding neighbourhoods) were asked about their attitudes to intensive housing, people fell into three camps with regard to attitudes to higher density housing: Acceptors people who think higher density living is OK, and are living that way now, or could see themselves living that way in the future. Rejecters people who think that higher density living is socially bad and should not be allowed, full stop. Intellectual acceptors people who accept the idea that higher density living may suit some people, but it s not for them. Page 45

50 While the research does not allow for an accurate fix on the number of people in these different categories, the numbers suggest an uneven split, with about 40% of people falling into the rejecter camp, and about 30% each in the acceptor and intellectual acceptor camps. One possible interpretation of this outcome is that the limit to intensification is 60% (i.e. all acceptors plus all those who are intellectually not opposed). For this reason, a scenario was modelled on the base run, with an intensification factor of 60% incorporated in the allocation process 6. In the first instance household demand is allocated to dwellings to the point where capacity becomes constrained. Once supply cannot meet first choice preferences, re allocation in each time period takes place in the following steps: Step 1: demand allocated across density categories as per preferences, Step 2: residual demand for conventional dwellings (unallocated in step 1); allocated to medium density dwellings, Step 3: residual demand for conventional dwellings (unallocated in step 2); allocated to high density dwellings Step 4: demand for medium density dwellings; allocated to medium density dwellings Step 5: residual demand for medium density dwellings (unallocated in step 4); allocated to high density dwellings Step 6: demand for high density dwellings; allocated to high density dwellings Within each step, allocation takes place in successive rounds. In the first round, if capacity exceeds demand, all demand is accommodated in the TA of choice. Otherwise, demand is allocated as capacity dictates. Following this initial allocation, capacity is recalculated for each TA. The process is summarised in Figure 5.1, below. In the second round, demand for the specific dwelling type is accommodated in areas other than the TA of choice. A migration factor is calculated, based on the remaining capacity in each TA and Census 2006 data for households TA of usual residence and TA of usual residence five years ago (Table 5.4). The data in Table 5.4 shows, for example, that of the households that moved from Rodney District within Auckland Region, 39% moved to North Shore City. Revised demand is calculated based on the migration factor and is allocated in the same manner as the initial demand in the first round. 6 Building a Better Future: Intensification Review (2000) Research Solutions. Page 46

51 Figure 5.1: Growth Allocation Process Summary SUPPLY CONVENTIONAL MEDIUM HIGH DEMAND CONVENTIONAL MEDIUM STEP 1 Round 1: TA of choice Rounds 2 7: Other TAs STEP 2 Round 1: TA of choice Rounds 2 7: Other TAs STEP 4 Round 1: TA of choice Rounds 2 7: Other TAs STEP 3 Round 1: TA of choice Rounds 2 7: Other TAs STEP 5 Round 1: TA of choice Rounds 2 7: Other TAs HIGH A B I L I T Y T O STEP 6 P A Y HIGH Round 1: TA of choice Rounds 2 7: Other TAs LOW Table 5.4: Household Migration by TA of Usual Residence and TA of Usual Residence 5 Years Ago, 2006 UR 5 Years Ago UR 2006 RDC NSC WCC ACC MCC PDC FDC RDC NSC WCC ACC MCC PDC FDC Subsequent rounds are carried out in the same manner as the second round until either all demand has been accommodated or all capacity has been exhausted. If unused residual capacity exists at the end of one time period, it is carried over into the following time period. In other words, growth is delayed until the point where additional capacity comes on line. Page 47

52 Local Area Allocation Following the TA level allocation, the AU level allocation is calculated. In this instance, it is only demand that has already been allocated at the TA level that is subsequently allocated to those area units with capacity for development within the TA (i.e. there is no longer the case where demand within the TA exceeds capacity within the TA). The AU level allocation uses an established suburbs method that assumes households seeking a specific dwelling type will display a preference for area units with an established supply of that dwelling type. A demand factor is calculated, based on the areas with capacity and the number of established dwellings (by dwelling type) that already exists in each area. In the first round, if capacity exceeds demand, all demand is accommodated in the AU of choice. Otherwise, demand is allocated as capacity dictates. Following this initial allocation, capacity is recalculated for each AU. Four rounds are carried out in this manner (sufficient to allocate over 99% of demand in all instances). A final fifth round is calculated solely on the basis of residual capacity, in order to allocate demand in those few instances where capacity has been created in an area that previously had no dwellings. 5.3 Results: TA Summary Outcomes Results from the Base Run supply scenario modelling are presented below. Results from other runs primarily carried out to test the sensitivity of assumptions are reported in summary here and in detail in Appendix A Medium Household Growth Scenario: Base Run Supply Scenario Under the medium household growth future and supply assumptions of the Base Run scenario (all rural capacity available, general infill consuming 80% of available infill space and redevelopment 20%) the region runs out of capacity in The key constraints begin to manifest in North Shore City, Auckland City and Manukau City in the period 2012 to 2016, when the supply of conventional residential land is exhausted (Table 5.6). By the end of 2021, there are considerable constraints on household choice for conventional density dwellings in particular, in all TAs except Rodney (due to the large volume of rural capacity contained within this scenario). Under the Base Run scenario, the region does not have sufficient capacity to accommodate all growth to 2031, regardless of how the demand is allocated. In the final period , additional demand for dwellings is 45,951 however the model is only able to allocate 24,030 a shortfall of almost 22,000 units. Page 48

53 This means that figures in Table 5.5 do not match actual household forecasts to 2031, rather they reflect the results of the modelled allocation. A full set of tabular outcomes for the High household growth scenario along with the other supply scenarios are appended to this report. Alternative demand scenarios have also been run to test an intensification first situation described in the previous section, the results of which are summarised below with detailed outputs also contained within the Appendix A Table 5.5: Allocation of Households to Total Residential Capacity RDC NSC WCC ACC MCC PDC FDC ARC Incremental Capacity 1,835 2,263 1, ,983 2, ,354 Total Capacity 28,060 22,368 21,580 57,692 27,541 5,984 5, ,315 Demand for Dwellings 2,067 3,590 3,542 7,981 6, ,369 Dwellings Allocated 2,082 3,607 3,556 8,040 6, ,369 Total Dwellings 36,115 77,422 67, , ,982 15,898 15, ,108 Residual Capacity 25,978 18,761 18,024 49,651 20,904 5,221 4, ,946 Incremental Capacity 3,118 1,913 1,997 1,250 4, ,195 Total Capacity 29,097 20,673 20,021 50,901 25,679 5,363 4, ,141 Demand for Dwellings 3,689 6,495 6,183 14,371 11,970 1,250 1,630 45,587 Dwellings Allocated 4,310 6,495 7,411 12,459 11,021 2,229 1,662 45,587 Total Dwellings 40,425 83,917 74, , ,003 18,126 17, ,695 Residual Capacity 24,786 14,178 12,610 38,442 14,658 3,134 2, ,553 Incremental Capacity 1,286 1,157 2,063 1,125 1, ,757 Total Capacity 26,072 15,335 14,673 39,567 15,783 3,135 2, ,310 Demand for Dwellings 3,960 6,552 6,387 14,331 12,503 1,305 1,739 46,777 Dwellings Allocated 14,870 5,371 7,967 4,989 9,184 2,846 1,551 46,777 Total Dwellings 55,295 89,288 82, , ,186 20,972 19, ,472 Residual Capacity 11,202 9,964 6,707 34,578 6, ,193 70,533 Incremental Capacity Total Capacity 11,202 9,964 7,129 34,578 6, ,193 70,956 Demand for Dwellings 3,989 6,561 6,424 14,262 12,681 1,292 1,717 46,926 Dwellings Allocated 11,169 6,354 5,834 15,487 6, ,193 46,926 Total Dwellings 66,465 95,641 88, , ,786 21,261 20, ,398 Residual Capacity 33 3,611 1,295 19, ,030 Incremental Capacity Total Capacity 33 3,611 1,295 19, ,030 Demand for Dwellings 3,739 6,331 6,320 13,809 12,815 1,313 1,624 45,951 Dwellings Allocated 33 3,611 1,295 19, ,030 Total Dwellings 66,497 99,252 89, , ,786 21,261 20, ,428 Residual Capacity Table Row Headings Incremental Capacity: Total Capacity: Demand for Dwellings: Dwellings Allocated: Total Dwellings: Residual Capacity: period Capacity added in time period Total available for allocation Household growth anticipated in time period Number of households able to be placed during time period Total households at end of period in each TA and Region Capacity left at end of period available for households in next Page 49

54 Table 5.6: Allocation of Households to Conventional Residential Capacity RDC NSC WCC ACC MCC PDC FDC ARC Incremental Capacity , ,228 2, ,062 Total Capacity 21,676 4,621 8,952 8,387 9,964 4,387 3,712 61,698 Demand for Dwellings 1,387 1,896 2,233 3,217 4, ,147 Dwellings Allocated 1,387 1,896 2,233 3,217 4, ,147 Total Dwellings 31,123 57,598 55,379 92,868 83,495 13,167 14, ,903 Residual Capacity 20,288 2,725 6,719 5,170 5,621 3,968 3,060 47,551 Incremental Capacity 1, , ,439 Total Capacity 21,692 2,924 8,413 5,170 5,621 4,110 3,060 50,990 Demand for Dwellings 2,313 3,184 3,543 5,868 7, ,043 23,883 Dwellings Allocated 2,793 2,924 4,352 5,170 5,621 1,430 1,593 23,883 Total Dwellings 33,915 60,522 59,731 98,038 89,116 14,597 15, ,786 Residual Capacity 18, , ,680 1,466 27,107 Incremental Capacity , ,169 Total Capacity 19, , ,681 1,466 29,276 Demand for Dwellings 2,378 2,750 3,362 5,177 6, ,036 22,196 Dwellings Allocated 12, , ,681 1,466 22,196 Total Dwellings 46,313 60,654 65,250 98,038 89,116 17,278 17, ,982 Residual Capacity 7, ,080 Incremental Capacity Total Capacity 7, ,080 Demand for Dwellings 2,262 2,459 3,093 4,958 6, ,822 Dwellings Allocated 7, , 080 Total Dwellings 53,393 60,654 65,250 98,038 89,116 17,278 17, ,062 Residual Capacity Incremental Capacity Total Capacity Demand for Dwellings 2,113 2,485 3,147 4,833 6, ,974 Dwellings Allocated Total Dwellings 53,393 60,654 65,250 98,038 89,116 17,278 17, ,062 Residual Capacity In order to ensure the model accommodates households appropriately residual capacity in a given period is added to new capacity as it comes on line to provide a total capacity able to accept household growth in the new period. In Table 5.6 this is shown by the 2011 remaining supply figure for conventional housing stock of 47,551 becoming 50,990 in It is important to also note that the large capacity to accommodate growth in Rodney District is primarily rural. In order for this to be utilised, significant changes to the manner in which people live will be required. Note also, that as capacity is exhausted elsewhere, Rodney is modelled to accommodate more of the region s growth in primarily rural locations. This is restricted in scenarios tested below. Page 50

55 Table 5.7: Allocation of Households to Medium Residential Capacity RDC NSC WCC ACC MCC PDC FDC ARC Incremental Capacity , ,575 Total Capacity 3,154 5,708 4,125 4,676 10,289 1, ,989 Demand for Dwellings ,998 1, ,053 Dwellings Allocated ,057 1, ,053 Total Dwellings 2,943 9,650 7,305 26,816 14,590 2,279 1,350 64,933 Residual Capacity 2,738 4,883 3,310 2,619 8, ,936 Incremental Capacity 1, , ,005 Total Capacity 4,141 5,709 3,310 2,619 11, ,941 Demand for Dwellings 816 1,596 1,615 3,833 3, ,179 Dwellings Allocated 957 1,856 2,035 2,619 4, ,179 Total Dwellings 3,900 11,506 9,340 29,435 18,606 2,975 1,350 77,112 Residual Capacity 3,184 3,853 1, , ,762 Incremental Capacity , ,253 Total Capacity 3,763 4,402 1, , ,015 Demand for Dwellings 936 1,829 1,853 4,164 3, ,918 Dwellings Allocated 1,826 3,265 1, , ,918 Total Dwellings 5,726 14,772 10,616 29,435 26,111 3,021 1,350 91,030 Residual Capacity 1,937 1, , ,097 Incremental Capacity Total Capacity 1,937 1, , ,097 Demand for Dwellings 1,032 2,008 2,078 4,343 4, ,076 Dwellings Allocated 1,937 1, , ,097 Total Dwellings 7,663 15,908 10,616 29,435 27,134 3,021 1,350 95,127 Residual Capacity Incremental Capacity Total Capacity Demand for Dwellings 981 1,900 1,991 4,183 4, ,490 Dwellings Allocated Total Dwellings 7,663 15,908 10,616 29,435 27,134 3,021 1,350 95,127 Residual Capacity Tables 5.6, 5.7, and 5.8 present the allocation for conventional density, medium density, and high density for each TA under the medium growth scenario. It is important to note that as conventional and then medium density capacity fills up households are forced into high density. This means that the figures being allocated in each of the later time periods are higher than the amount of demand for dwellings that naturally arises. Page 51

56 Table 5.8: Allocation of Households to High Residential Capacity RDC NSC WCC ACC MCC PDC FDC ARC Incremental Capacity ,717 Total Capacity 3,230 12,039 8,504 44,628 7, ,379 77,627 Demand for Dwellings , ,169 Dwellings Allocated , ,169 Total Dwellings 2,049 10,174 4,492 34,998 5, ,271 Residual Capacity 2,952 11,153 7,996 41,862 6, ,347 72,459 Incremental Capacity ,250 2, ,751 Total Capacity 3,264 12,040 8,298 43,112 8, ,347 77,210 Demand for Dwellings 561 1,715 1,025 4,670 1, ,525 Dwellings Allocated 561 1,715 1,025 4,670 1, ,525 Total Dwellings 2,610 11,889 5,517 39,668 7, ,797 Residual Capacity 2,703 10,325 7,273 38,442 7, ,278 67,684 Incremental Capacity , ,335 Total Capacity 2,832 10,801 7,878 39,567 7, ,278 70,019 Demand for Dwellings 646 1,973 1,172 4,989 1, ,663 Dwellings Allocated 646 1,973 1,172 4,989 1, ,663 Total Dwellings 3,256 13,862 6,689 44,657 8, ,459 Residual Capacity 2,185 8,828 6,707 34,578 5, ,193 59,357 Incremental Capacity Total Capacity 2,185 8,828 7,129 34,578 5, ,193 59,779 Demand for Dwellings 695 2,094 1,253 4,960 1, ,029 Dwellings Allocated 2,153 5,217 5,834 15,487 5, ,193 35,750 Total Dwellings 5,409 19,079 12,523 60,144 14, , ,209 Residual Capacity 33 3,611 1,295 19, ,030 Incremental Capacity Total Capacity 33 3,611 1,295 19, ,030 Demand for Dwellings 645 1,946 1,183 4,793 1, ,486 Dwellings Allocated 33 3,611 1,295 19, ,030 Total Dwellings 5,441 22,690 13,818 79,236 14, , ,238 Residual Capacity Page 52

57 5.4 Results: Years till Exhaustion A key metric in helping develop understanding of adequacy is years till exhaustion. While it is possible to assess each individual TA by analysing the figures in Tables 5.5 to 5.8, for convenience they are summarised in Table 5.9. Conventional capacity in North Shore City is exhausted in 2016, in Auckland City, 2016 and in Manukau City in 2015 (Table 5.9). Table 5.9: Years till Total ARC Capacity Exhaustion by Density Type. Territorial Authority Conventional Medium High Total Rodney District North Shore City Waitakere City Auckland City Manukau City Papakura District Franklin District Auckland Region At the regional level and in total, conventional capacity will be exhausted by Note that the Base Run includes all rural capacity (some 30,570 units). Medium density capacity is exhausted in 2023 and high density in These timings are influenced to a greater, or lesser, degree by unallocated demand from lower density classes (conventional to medium and medium to high). Auckland s local authorities are well aware of the shortages outlined in the analysis above and have a range of projects and processes in place to address (as much as possible) the shortfalls. In general this action takes two forms. Firstly, councils are identifying additional areas around public transport nodes and town centres that are able to be intensified and/or redeveloped to provide additional residential capacity. This will raise capacity overall although potentially not for conventional density necessarily. Secondly, councils are engaged in development of the social infrastructure in and around medium and high density centres. This will, in the medium to long term, make these areas more acceptable to families and those willing to trade additional private space for public space and public amenity. However, the mismatch between demand and supply is very significant. If nothing else changes, in less than 8 years households in North Shore City, Auckland City and Manukau City will have no options for standalone house living, the traditional New Zealand dwelling of choice. This timing falls well within the current LTCCP planning horizons (10 years) and is going to require considerable effort on the part of councils to address. Page 53

58 Council efforts are further hampered by the claimed lack of tools available to them to achieve the intensification objectives 7. As an example, councils currently do not have the ability to effectively aggregate land for large scale intensive developments as happens in many international cities. There is some evidence Councils are testing the waters in this regard, with Auckland City purchasing some strategic sites in Panmure. There is a sense that the tools available to Councils allow enabling of development but are not sufficient to allow directing of development. It should also be noted that some areas have a significant proportion of unoccupied dwellings, primarily holiday homes that have not been incorporated into the estimates of years to exhaustion. It is true that as areas mature, especially those close to the metropolitan centres, holiday homes get converted into permanently occupied dwellings. This was the case on the Hibiscus Coast, now heavily populated by permanent residents, whereas years ago was primarily a holiday destination for Aucklanders. Discussion with Rodney District officials revealed that some 16% of all dwellings in Rodney are unoccupied (some 6,300). In the long run, this could add a couple of years to the conventional capacity for the region, however it is unlikely that the overall percentage of unoccupied would drop to zero. Rather new areas will come on stream as holiday destinations, once current areas fill up. In new development areas, unoccupied homes represent recent development not yet sold or moved into rather than holiday homes. This is the case in and around the Flatbush structure plan area, where the 2006 census counted some 10% of the dwellings either unoccupied or under construction (5.9% and 4.2% respectively). For the purposes of this report, these are not considered part of the future capacity, as the majority are occupied within 3 months (although this time may lengthen under the current housing market slow down). It is noted that Councils may view this stock as future capacity and could lead to a difference in estimates of future residential capacity. The Flatbush figure is higher in total than the regional average of 7.9% (7.1% unoccupied and 0.9% under construction) due to the high level of under construction dwellings Sensitivity Testing Under a high growth future capacity is expected to become exhausted in 2025, region wide, or some 3 years earlier than the medium growth future. Conventional density dwellings are expected to run out a couple of years earlier in North Shore, Auckland and Manukau (2014), with Medium density in Auckland City running out in the same year (Table 5.10). 7 As reported in the RGS Evaluation by the local authorities Page 54

59 Table 5.10: Years till Capacity Exhaustion by Density Type: Base Run, high growth. Territorial Authority Conventional Medium High Total Rodney District North Shore City Waitakere City Auckland City Manukau City Papakura District Franklin District Auckland Region An alternative scenario has been run that allows for households to seek higher density dwellings in order to remain within their chosen area rather than shifting (Table 5.11). In this scenario, capacity is extended by 2 years over the base run in total, however, this is driven by the extension to run out time in Rodney. In the main metropolitan TAs conventional capacity runs out within the next 8 years. This implies that even under an optimistic adoption of higher density forms of housing (60%), capacity under the Base Run assumptions is highly constrained. Table 5.11: Years till Capacity Exhaustion by Density Type: Intensification, Medium Growth. Territorial Authority Conventional Medium High Total Rodney District North Shore City Waitakere City Auckland City Manukau City Papakura District Franklin District Auckland Region The next two scenarios present situations where rural capacity is firstly constrained to 50% of current Countryside Living and all rural towns and, secondly to only rural towns. General rural capacity is dropped in both. Page 55

60 Table 5.12: Years till Capacity Exhaustion by Density Type: Restricted Rural, Medium Growth. Territorial Authority Conventional Medium High Total Rodney District North Shore City Waitakere City Auckland City Manukau City Papakura District Franklin District Auckland Region Table 5.13: Years till Capacity Exhaustion by Density Type: No Rural, Medium Growth. Territorial Authority Conventional Medium High Total Rodney District North Shore City Waitakere City Auckland City Manukau City Papakura District Franklin District Auckland Region Under these scenarios, conventional capacity runs out in 2020, however in most TAs conventional capacity runs out in 6 7 years. These scenarios, or ones very similar, are likely to represent the closest approximation to Auckland s residential future. Rural capacity is not a realistic option for accommodating Auckland s residential growth; it does not contribute to a Compact City form and Countryside living is likely to be beyond the financial reach of most households. Note that under a higher growth future the capacity runs out faster than presented here, which means the risks faced by Auckland s Councils are significant indeed. The final scenario run explores the impact of 50% of infill opportunity being subject to the higher densities of redevelopment (compared with 20% in the Base Run). Page 56

61 Table 5.14: Years till Capacity Exhaustion by Density Type: High Redevelopment, Medium Growth. Territorial Authority Conventional Medium High Total Rodney District North Shore City Waitakere City Auckland City Manukau City Papakura District Franklin District Auckland Region The High redevelopment scenario adds some 10,000 additional dwellings to capacity split between North Shore, Waitakere, Auckland and Manukau Cities. This represents little more than one year s demand growth. It is likely that alternative measures including significant re zoning and potentially incentives to developers to maximise zonings or even exceed them (where appropriate). 5.5 Results: Suburb Area Outcomes At the suburb level the variations in supply are more marked. Map 5.1 outlines the distribution of residential capacity for Areas that are Red have the least supply, areas of green the most. Areas that appear blank have no capacity remaining. Note that Map 5.1 is a snapshot of available capacity as at 2008, therefore it does not include the majority of the Structure Plan Areas where capacity is planned to come on stream in future years. Page 57

62 Map 5.1: Auckland Regional Residential Capacity as at June 30, 2008 Page 58

63 By 2021 this situation has changed dramatically under the Base Run scenario assuming no other capacity becomes available other than that scheduled to come on line under current zoning and structure plan timelines. Map 5.2: Auckland Regional Residual Residential Capacity, 2021 Page 59