Housing market areas in England: change and stability 2001 -2011 Mike Coombes CURDS, Newcastle University acknowledgements: Colin Wymer (CURDS) Census datasets are Crown Copyright
The significance of the definition of Housing Market Areas Legislation requires planning authorities to work within housing market area (HMA) geography when assessing housing need, co-operating with adjacent authorities where HMAs straddle their boundaries Despite this imperative there are no ‘official’ HMA definitions to determine which authorities should co-operate, and the minimal guidance refers only to the relevance of three different datasets – commuting, migration, house prices – and other functional economic areas such as Travel to Work Areas (TTWAs) The now-defunct National Housing and Planning Advisory Unit (NHPAU) had funded research to devise ‘state-of-the-art’ HMA definitions to supplant the inconsistently defined HMAs had already had to be adopted in all areas The advent of the 2011 Census commuting and migration data has led to the new research reported in this paper, which uses the NHPAU methods to assess how far dramatic changes in the housing market over the decade 2001 -2011 have changed the geography of HMAs Whether an area is deemed to have a substantial need for new housing will depend on the areas covered by the housing need calculations: an area may be able to maintain a ‘NIMBY’ position if the HMA definitions do not link it with any adjacent areas where housing shortages are critical
Background: earlier work in CURDS defining HMAs Consortium led by Heriot-Watt University were funded by NHPAU to define HMAs because ‘SHMAs’ identified in the preceding years were not comparable… ! different methods ! methods applied inconsistently ! often appeared to have set out to show LAs were each self-contained, thereby minimising need for the LAs to closely co-operate ! widely differing scales of HMAs (some were ‘strategic’ but some ‘local’)… eg. Cumbria in the NW split into 9 S-RHMAs, while the whole of the W. Midlands split into a mere 4 (and looking at their boundaries, shown on the right, makes it clear that they aren’t functional economic areas but instead are grouping urban with urban, and rural with rural)
CURDS research devised new methods to be applied consistently everywhere and taking account of the datasets seen as relevant in defining HMAs: > commuting flows > migration flows > [and as a check] house prices The basic issues in any regionalisation led to 9 guidelines for definitions of HMAs: What area must the set of HMAs to be defined cover? analyses should cover the whole country to avoid ‘edge effects’ Can the HMAs overlap? this is not a practical option, but a two-tier set of HMAs offers some insights Will centres be defined for each HMA? analyses should not pre-determine HMA centres but accept all ‘clusters’ of flows Will the HMAs be contiguous areas? analyses should not impose this to avoid ‘sub-optimality’ (it is ‘built in’ the data) How many HMAs will be defined? this depends on the data and choices made for ‘parameters’ in the analyses Which data will be analysed? commuting and migration need considering; house price data is supplemental What zones will the HMAs be built from? wards as ‘frozen’ for the release of data from the 2011 Population Census Will the HMAs be groups of LAs? ‘gold standard’ results are groups of wards, then there are ‘best-fits’ to whole LAs How will the analyses be carried out? the method now established as the most effective is that which defines TTWAs…
Dividing a territory into areas: the TTWA method This is an internationally recognised model for labour market area definitions The key features of the TTWA method are instrumental for defining areas Ø all ‘building block’ zones to be in (only) 1 TTWA: no omissions or overlaps Ø group the zones 1 -by-1: not a top down process Ø no ‘centres’ so no catchment areas: polycentric TTWAs emerge if they exist Ø no contiguity constraint (nb. non-contiguous groupings are in fact rare) The most important measure is area “self-containment” in terms of flows… eg. in commuting analyses: % of jobs in the area held by local residents AND % of employed residents working in the area Can set different self-containment criteria but the method combines these with the concern to avoid very small areas because they can have unreliable trends, with some ‘trade off’ between criteria:
Dividing territories into areas: the TTWA method Brief visualisation excerpt from an analysis: 6 maps show state-of-play at points in the process to sequentially identify and disband the ‘proto’ TTWA that is furthest from meeting the set criteria, reassigning its zones The maps show the ‘proto’ TTWAs as at the point when in the whole UK there were: 8, 000 4, 000 2, 000 1, 000 500 & 250 (by when nearly all meet the final criteria) On these maps (showing central southern England, from an analysis of the UK): the blue boundaries are the final 2001 -based TTWAs the green boundaries are Local Authority areas the red ‘stars’ connect the LSOAs grouped together at that step of the process [+ selected towns/cities shown with 2 letter codes] The sequence of maps gives a brief ‘movie’ of the process…
8000 ‘proto’TTWAs
4000 ‘proto’TTWAs
2000 ‘proto’TTWAs
1000 ‘proto’TTWAs
500 ‘proto’TTWAs
250 ‘proto’TTWAs
Key results from 2001 Census data analyses Commuting distance varies little by region (but more between the urban and rural areas which tend to be grouped by functional economic areas) …whereas migration distances vary markedly by broad region of the country km
Approach to analysing the 2011 Census Guidelines for the NHPAU work were followed to ‘update’ it so far as possible: Ø HMAs meet the same self-containment levels for commuting and migration flows Ø both the 2 -tier and the ‘unitary’ NHPAU versions of HMAs have been produced Ø HOWEVER there are inevitable changes, including altered wards in some areas Ø 2011 migration data analysed with(out) those aged <25 (cf. data for NHPAU) Ø some data on Scotland unavailable so there may be ‘border effects’ on the results Ø assumed that the minimal influence of house prices on the 2001 -based HMA definitions for NHPAU meant that this extra step could be dropped BASIC RESULTS in terms of numbers of HMAs that cover England & Wales: Strategic (of 2 tier) Unitary 2001 -based (NHPAU) 84 103 2011 -based (nearest equivalent) 78 93
‘Updating’ Strategic HMAs with data on 2010 -1/2011 2000 -1/2001 2010 -1/2011 There is greater stability/inertia to the Strategic than the Local scale of HMAs… however judging how ‘surprising’ a level of change is not simple (also the datasets available from the 2 Censuses were not identical, and the effect of this is not measurable … although it is probably not major? )
Summarising the implications: HMAs and policy Looking at the recent (2006 -13) change in house prices shows a ‘ripple’ from London that is very localised There are echoes of this in some of the changes to HMAs: these are greatest south and west of the capital (eg. western Hants. ) and also around Cambridge HOWEVER it must be remembered that changes to HMAs can be at least partly due to the analysis method that was used, and the specific criteria applied Basic conclusions on ‘updating’ the definition of HMAs: There is a similar level of stability to the boundaries as seen with TTWAs for decades: * most are broadly unchanged * there are some ‘transfers’ between areas * there are some ‘merges’ * there are very few ‘de-merges’ …this does not amount to a strong case for updating the current S-RHMAs BUT the current S-RHMAs are so inconsistently defined they must be causing a very uneven planning for much-needed additional housing SO the new 2011 Census datasets create the opportunity to rigorously define a new set of HMAs, identifying which LAs need to co-operate to deliver new housing
Скачать презентацию Housing market areas in England change and stability
5ea57227f00b872dda52e225e587f64e.ppt