Strategic Housing Market Assessment: A Case Study from England

The first case study we present here is drawn from the UK, and specifically from England. As argued in Chap. 4, England is an interesting case study both because it has a relatively large social housing sector and because over recent years it has developed a significant element of affordable housing requirements being built into the land use planning system. Practices for planning overall housing growth numbers/locations and for planning and resourcing social/affordable housing provision have been explicitly brought together under the banner of ‘Strategic Housing Market Assessment’ (SHMA) since the mid-2000s. The case study chosen is the city region centred on Bristol, in the South West of England, referred to here as ‘Wider Bristol’, set within a sub-region known as the ‘West of England’. One of the authors lived and worked in Bristol for two decades and undertook a number of studies of housing need and affordability for the local authorities, although the current assessment reviewed here (2015-based) was undertaken by consultants.

Wider Bristol sits within the part of the South West region which is relatively close to both London (less than 200km away) and the heart of England (see Map 10.1). As such it is a relatively prosperous and higher demand area in housing market terms, and could be described as being part of the ‘Greater South East’ (Bramley 2015a). It comprises three

Map 10.1 Greater Bristol (Source: Bramley 2015a)

unitary local authorities, with the core City of Bristol and two adjacent areas, North Somerset and South Gloucestershire, with a total population of 910,000 (384,000 households in 2015).

The geographical delineation of the HMA is the first substantive issue addressed by the SHMA. This issue has received considerable attention in the UK over recent years (Jones et al. 2010). The previous SHMA used a wider area for the ‘West of England’ HMA, including the neighbouring small city of Bath and its hinterland (including parts of Wiltshire and Somerset). However, the consultants undertaking the 2015 SHMA for the local authorities argue that Bath and its hinterland should be treated as a separate HMA. They are able to produce a range of evidence in terms of commuting flows, migration flows and house prices which appears to justify this stance in terms of criteria suggested in the official guidance. While the local authorities prefer this separation, some of the other groups involved, for example, the economic development partnership and the housing sector forum, have questioned it.

Central government provides guidance to local authorities on how they should carry out SHMAs (CLG 2007, 2014), which is quite prescriptive on some points while being light touch on other issues. There is a broad distinction between existing unmet housing needs and needs expected to arise in the future, over the plan period, and a division between needs for ‘affordable housing’ and needs/demands for market housing. The definition of affordable housing is crucial here. In England, until recently, one could say fairly confidently that affordable housing was housing to rent or buy for which the cost was significantly below market rates. In effect that meant subsidised housing, including public (council) and housing association (registered provider) ‘social’ rented housing, ‘intermediate’ or ‘affordable’ rented housing, or ‘low- cost home ownership’ (LCHO, typically shared ownership/equity). Whether people could afford home ownership was assessed by applying typical mortgage lending criteria to market prices towards the lower end (e.g. lower quartile, for relevant size of unit). As private renting has become more important, the application of equivalent ratio criteria (e.g. 25 % of gross income, 30-35 % of net income) to local market rents has become more the operative threshold. However, in this particular case study, the consultants used the proportion of households claiming Housing Benefit/Local Housing Allowance as the threshold to distinguish households in need of affordable housing, arguably too narrow a definition.

Table 10.1 summarises the part of the Wider Bristol 2015 SHMA which deals with existing unmet need for affordable housing. These needs are identified under four main headings: ‘homeless’; ‘concealed households’; ‘overcrowding’; and ‘other unsuitable’, all categories mentioned in the Department for Communities and Local Government (DCLG) (2014) practice guidance. It appears from the table that the largest of these, numerically, is overcrowding, which perhaps tallies with evidence in the report of a large rise in overcrowding in this area between the 2001 and 2011 Censuses. The homeless categories reflect the highly developed statutory framework around homelessness in the UK; in England, households found to be homeless (including at imminent risk of losing their home) and in priority need (e.g. with children, vulnerable elderly) have what is tantamount to a right to rehousing as well as advice and support.

Local authorities are obliged by law to compile and report data on how they fulfil their Statutory Homelessness duties and these data provide the source for first four items. Concealed households could be based on a combination of waiting list, national sample survey and census data; in practice in this case the number presented is an estimate of the increase in concealed families (couples and lone parents living with others) between the 2001 and 2011 Censuses. Overcrowding is based on the Censuses, adjusted (down) for consistency with the national English Housing Survey measure which uses the stricter bedroom standard (closer to DCLG Guidance). The categories of ‘others in unsuitable housing’ are based on local authority waiting list numbers in the relevant ‘Reasonable Preference’ categories (these ‘RP’ categories comprise: homeless; unsatisfactory housing; medical/welfare grounds; others needing to move to avoid hardship). The general approach entails not only fairly conservative need criteria but also an emphasis on avoiding double counting and not including households who could afford suitable housing in the market. It may be noted that excluded from these numbers

Table 10.1 Assessing current unmet gross need for affordable housing—Wider Bristol HMA 2014



Increase in

Components of unmet housing need

housing gross need

housing supply

housing need

Homeless households in priority need

Currently in temporary accommodation—communal



In temporary accommodation—market housing


In temporary accommodation—affordable housing (LA, HA)



Accepted as homeless but without temporary accommodation



Concealed households

Growth in concealed families aged <55



Overcrowding based on 'bedroom standard'

Overcrowded in private rented housing


Overcrowded in social rented housing



Others in unsuitable housing & can't afford market

Need to move for medical/welfare/disability reasons



Need to move locality to avoid hardship







Source: Wider Bristol 2015 SHMA are (a) those living in unsanitary or otherwise physically unsatisfactory housing (it being argued that the right solution to these problems is to improve the housing); (b) owner occupiers (who will typically have the equity to buy a suitable home); (c) students (who are transient); (d) concealed single person potential households, regardless of preferences and (e) single sharing households (for whom the government will now only subsidise a room in shared accommodation up to age 35).

The main need figures are shown in the first column, totalling 11,228. The figures in the second column highlight households who currently occupy a social rented dwelling, and therefore do not need/warrant a net addition to housing supply—they could move to a more suitable dwelling and release a vacancy for someone else. The third column highlights groups who are not currently counted in the number of households, and so represent an addition to housing requirements over and above any baseline household estimates or projection.

By way of commentary, the total gross need is 3.0 % of total household, which may be argued to be a conservative figure. For example, in a comprehensive model covering all HMAs in England a comparable need estimate of 7.2 % is shown for Wider Bristol (Bramley and Watkins 2016; Bramley 2015a). Differences are probably mainly due to the latter study including concealed singles who want to move, sharing households and some people with affordability/payment problems. The homeless data reflects national downward trends in recorded homelessness, but a national monitoring study (Fitzpatrick et al. 2015b) shows clearly that this is due to enhanced prevention activity, and the more reliable combined homeless activity including prevention cases shows continued rising demand, particularly in the south of England. The use of the Housing Benefit/Allowance (HB/LHA) entitlement as a proxy for affordable housing need threshold raises various problems—it ignores people who do not get HB but still have affordability problems on reasonable objective and subjective measures (Bramley 2012b), and the fact that obtaining HB/LHA does not necessarily prevent people experiencing such problems. Also, with this kind of measure, the more the British Government restricts or cuts HB/LHA, the lower the apparent level of need regardless of the objective situation.

The results from this part of the analysis are then combined with other elements in the overall assessment model to contribute to an overall need

Table 10.2 Assessing total need for market and affordable housing—Wider Bristol Housing HMA 2014-2036

Components of need










1. Unmet need for affordable housing 2014 (from Table 10.1)



2. Supply of housing vacated (from data underling Table 10.1)




3. Overall impact current affordable need (rows 1+2)




4. Net increase in affordable housing need 2014-16



5. Forecast affordable housing completions



6. Projected impact affordable housing needs 2016 (rows 3+4+5)




7. Newly forming households 2016-36




8. Household dissolutions following death




9. Net household growth within Wider Bristol HMA (rows 7+8)




10. Impact of existing households falling into need




11. Impact of existing households climbing out of need




12. Impact of households migrating to/from area




13. Future need for market and affordable housing 2016-36 (rows 9+10+11 + 12)




14. Projected impact of affordable housing need in 2016 (row 6)



+ 2111

15. Total need for market and affordable housing (rows 13+14)




16. Average annual need for housing




17. Proportion market/affordable %

63 %

37 %

100 %

Source: As for Table 10.1

figure, as shown in Table 10.2. This time the columns refer to market housing need/demand, affordable need and overall housing requirement, respectively. The baseline needs from Table 10.1 refer to 2014, and these need to be rolled forward to 2016, the base year for the land use plan. This adds a small amount to the need ‘backlog’ because newly arising need somewhat exceeds the known likely number of affordable housing completions. The next section of the table covers the most important driver of the overall need assessment, looking forward 20 years, which is the household projection. Net household growth of 89,300 is shown to be the result of207,600 new households forming, while 118,330 dissolve through death. The division of these numbers into those able to afford market housing and others relies upon the analysis of current/recent HB/ LHA claiming by households by age group and household type, linked to the household projections detailed outputs on household type by age at future dates. This assumes that the current benefit regime continues into the future, and that relative income distribution remains unchanged. The analysis uses the same underlying data to estimate (a) the number and proportion of newly forming households each year who need affordable housing; (b) the number and proportion of in-and-out migrant households likewise; and (c) the number of households whose circumstances and incomes change such that they move in or out of need. It can be seen that the dominant influence on future housing need is the large excess between new household formation and household dissolution, and the fact that a high proportion of newly forming households cannot, at the outset, afford market housing. The migration component, and the people falling into or climbing out of need components, are both net negatives; the former because migrants tend to be better off, and the latter because households tend to improve their capacity to afford housing as they mature (thanks to marriage/partnership formation, rising career incomes, acquisition of assets and equity).

Towards the bottom of the table, the existing unmet need is combined with the projected future need, and the resulting numbers shown as annual flows and percentage shares. In this instance, it is suggested that 37 % of the housing required over the plan period will need to be ‘affordable’.

Table 10.3 Full assessment of planning requirements for new housing—Wider Bristol HMA, 2016-36




1. Demographic starting point: DCLG household projections 2016-36



2. Adjustment for long-term migration trends



3. Baseline projections allowing for local circumstances (rows 1+2)



4. Adjustment for suppressed household formation rates



5. Baseline housing need based on demographics (rows 3+4+5)



6. Adjustment to balance jobs and workers


7. Adjustment in response to market signals


8. Adjustment in response to backlog of housing provision 2012-16


9. Combined impact of adjustments (higher of rows 7,8)


10. 'Full objectively assessed need for housing' 2016-36 (rows 5+9)


Source: Wider Bristol 2015 SHMA

The final stage in bringing the numbers together is shown in Table 10.3. Here, future household needs are translated into dwelling numbers, incorporating an allowance for vacancies and second homes. There is also an adjustment away from the official household projections, to reflect a view taken by the local authorities and the consultant that migration rates within the projection should be based on the previous ten years rather than the previous five. The most recent figures were influenced by a severe recession and a longer base period is argued to be more reliable. A second adjustment is to add the figure of 1421 from Table 10.1 (total of col. 3), household need which was suppressed/not revealed in the actual household numbers in the base year. Three further adjustments were then provided for, again as suggested in the DCLG practice guidance. On the first of these, jobs/worker balance, it is argued that no adjustment is required. On the second, a small adjustment is made on account of the fact that house prices were higher/market affordability poorer, than in benchmark areas or nationally—we comment further on this below. A third adjustment identified relates to the shortfall between affordable need and supply in the period 2012-16, but it should be noted that these adjustments are treated as overlapping rather than additive.

One of the authors has been involved in providing an independent commentary on and assessment of this particular SHMA, and some main points emerging from that are highlighted here, to draw out some of the more challenging underlying issues.

Firstly, as a general comment, the method and its results can be seen to be very strongly tied to and dominated by the traditional extrapolative demographic-based household projections. There are significant dangers in uncritical reliance on household projections (Bramley and Watkins 2014, 2016; Bramley 2015b), particularly around ‘circularity’ (restricted supply leading to low apparent future need) and getting out of phase with cycles (reducing planned provision when it should be increased). The extremely large reduction in the official projected growth for Wider Bristol between 2008 and 2012 projections undermines confidence, as do other features including the negative need attributable to migration. Second, the amount of attention given to market conditions and signals seems small and its impact on the results marginal. We can demonstrate from a reputable economic model of the system that the trade-off between supply increases and affordability improvements is roughly 5:1 in proportional terms, which means that the SHMA is quite wrong to suggest that a 7.5 % uplift in plan numbers is an appropriate response to market signals, when a more appropriate response would be in the range 35-60 %.

Thirdly, evidence can be presented that the employment growth potential of Wider Bristol is relatively high, among the highest of the comparable set of city-regions. In the light of this, it is surprising that the SHMA does not identify employment prospects as a factor weighing in favour of higher housing plans. Finally, the annual need for affordable housing (1432 or 37 % of all new supply), while probably justified in needs terms, lacks credibility as something which can be delivered in the actual policy and market situation which now prevails. Realistic affordable housing quotas in plans and s.106 agreements are likely to be lower, which means the only route to delivering this level of affordable housing will be through increasing the total housing numbers substantially.

< Prev   CONTENTS   Source   Next >