The Scottish Government

Listen

ISBN 0 7559 4925 0

This document is also available in pdf format (224k)

Introduction

1. The aim of this paper is to present a commentary on the implications of the evidence base for the preparation by councils to meet the 2012 commitment. The commentary draws on councils' responses to the 2012 preformed exercise, modelling of possible scenarios and the most recent homelessness statistics - published on 1 November 2005.

2. In assessing the implications of the 2012 commitment and planning to meet this councils and the Executive need to assess how both the demand for homelessness services will change over the period and the implications for homelessness services which will be required to meet this need. A particular concern is the possible mis-match in some council areas between the likely available supply of new social housing lets and the number of homeless households with a need for social housing.

The proforma exercise and Scottish Executive modelling

3. The 2012 proforma exercise - reported in Analysis of Local Authorities Proformas to assess their ability to meet the abolition of the Priority Need Test - aimed to draw out councils' assessments of the possible impact of key determinants of demand and supply. In assessing the proforma returns it is important to note that councils have taken widely different views about trends in, and the impact of, factors affecting both projected numbers of homeless applicants and the supply of social lets.

4. To provide a context for reviewing the implications of councils assessments in the 2012 proforma the Executive, with the assistance of Tribal HCH, has developed trend based models designed to illustrate future scenarios on a common sets of assumptions. This paper presents a summary review of councils assessments drawing on the trend based models and the most recent homelessness outturn figures for 2004-05.

Projected demand for permanent housing

Number assessed as homeless

5. Throughout the 1990s, out of about 40,000 applications a year councils assessed around 30,000 as homeless. Between 2000-01 and 2002-03 in the lead up to and following changes in local authorities' duties to single homeless applicants, there was significant increase in numbers of applicants for assistance and since 2002-03 there have been around 40,000 households assessed as homeless each year. Almost all of the increase in homeless applications over this period has been in numbers of single people. (Chart 1)

6. One of the most important areas of uncertainty in planning to meet the 2012 commitment is whether, and if so by how much, the number of homeless applicants will continue to increase between now and 2012. We believe that the increase in single homeless applicants between 2000-01 and 2003-04 was driven in part by the increased duties which councils now have to provide those single applicants who need it with temporary accommodation and advice and assistance. Part of the cause may also be related to the growth more generally in the number of single households in the population as a whole.

7. For the 2012 proforma exercise councils were offered projections of numbers of homeless households in 2007-08 and 2011-12 by household type and age, based on the assumption that the proportion of households of each type who are homeless remained the same over the period as in 2003-04. Because household projections show the number of single person households increasing over the period, the overall effect of this approach is to project increases in the number of single homeless in most council areas, and little change in numbers of other types of homeless household. The projection made no assumptions about the future impact of prevention activity. Councils were asked to use these demographic trend based projections as a starting point, to adjust them on the basis of their own assessment of likely trends and also the likely impact of future prevention activity.

8. Only one council - Glasgow - projected a significant reduction in applications as a result of improved prevention activity. Overall 14 councils projected higher levels of homelessness applications than implied by demographic trends and of these 7 councils projected levels of homelessness applications more than 50% higher. For those councils other than Glasgow which projected lower increases than implied by demographic trends projected levels of homelessness by 2011-12 were between 1% and 20% lower than the trend based projection.

9. Charts 2a - 2d compare outturn and projections for 4 groups of councils - mainly rural councils, defined for this analysis as councils with 50 or fewer persons per hectare, non-rural councils excluding the 4 cities, Glasgow, and the remaining 3 cities. Trend based projections based on 2004-05 outturn have also been added for comparison. For each of the groupings of councils, other than Glasgow, projections in the proformas are higher than trend based projections. The main features of the charts are:-

• The mainly rural grouping of councils experienced the highest rate of growth in homelessness between 2001-02 and 2003-04. Overall the proforma projection is about 5% higher than the trend projection.
• The non-rural grouping of councils (excluding the 4 cities) also experienced an increase in homelessness between 2001-02 and 2003-04, but at a lower rate than the rural grouping. For this group of councils the proforma projections imply an overall continuation of the rate of growth experienced following the introduction of the new duties. The demographic trend based projections - both 03-04 and 04-05 based suggest a lower rate of growth. Overall the proforma based projection for this group of councils is some 25% higher than the trend projection. 04-05 outturn is very close to 03-04 trend projection and significantly lower than proforma base.
• While Glasgow experienced some growth in homelessness between 2000-01 and 2001-02 the number of homeless applicants subsequently fell slightly to 2003-04 and has dropped dramatically between 2003-04 and 2004-05. Glasgow's proforma projection for 2011-12 is 25% below the 2004-05 trend projection.
• The 3 remaining cities experienced some increase in homelessness between 2001-02 and 2003-04, falling back to about late 1990s levels in 2004-05. For this group of councils combined the proforma projection lies between the 03-04 and 04-05 trend based projections. The proforma based projections for these councils are about 5% higher than the 2004-05 based projection and about 9% below the 2003-04 based projection.

Chart 2a:

Chart 2b:

Chart 2c:

Chart 2d:

Number in priority need

10. The effect of removing the priority distinction by 2012 is equivalent to moving to 100% of homeless households being in priority need by that date. From the mid 1990s the percentage of homeless applicants assessed as in priority need has been increasing from 55% in 1996-97 to 74% in 2004-05. To achieve a steady rate of movement towards 100% assessed as priority by 2011-12 would require the proportion of homeless applicants in priority need to increase by about 4 percentage points each year. This is a significantly higher rate than in the most recent period. (See chart 3).

Chart 3:

11. Single people form the main group of homeless applicants currently assessed as not in priority need. From chart 4 the main group of applicants currently not assessed as in priority need are single applicants aged between 18 and retirement with the proportion not in priority need falling from around 46% of homeless aged 18-24 to 30% of those aged 55-69. The very low proportions of those aged under 18 and over 60 assessed as non-priority reflects the vulnerability of the very young and older homeless applicants.

Chart 4:

12. Chart 5 shows for each of the age groups the number of single people assessed as homeless and the number of these assessed as priority and non-priority. 18-24 year olds are the largest group of both single homeless and non-priority homeless. However there are also significant numbers of non-priority homeless in the other age groups.

Chart 5:

13. Councils' proforma returns and the Executive's trend based modelling point to increases in numbers of homeless applications being mainly in the number of single person households - particularly those aged 16-24. Meeting the housing and support needs of these groups of single applicants will be an important issue in addressing the 2012 commitment.

The housing needs of those in priority need

14. Not all households assessed as in priority need at present secure a social let. Of the 28,000 priority applications actioned in 2004-05 some 16,000 (57%) were offered permanent accommodation, just under 2,000 ( 7%) were offered temporary accommodation, councils lost contact with a little under 6,000 (21%) and a little over 4,000 (15%) had other outcomes. Chart 6 shows the 2004-05 outcome for households in priority need by broad household type. The main point to note is that non-housing outcomes are a much higher proportion of the outcomes for single homeless applicants in priority need than for the other household types.

Chart 6: Households in priority need in 2004-05 by outcome of application

15. In projecting the need for permanent housing arising out of the 2012 commitment a key uncertainty is likely changes in the proportion of applicants in priority need who will not require social housing. In particular, how will the number of applicants who lose contact with the council before completion of their application change? In 2012 proforma returns and in the Executive modelling the assumption is made that the proportion who lose contact or who have other outcomes will fall. In general, 2012 projections suggest proportions for single people similar to those currently experienced by single parents and other households with children.

The overall assessment of increased need for permanent lets

16. Charts 7 a-d show the projected need for permanent lets for priority homeless applicants over the period to 2012 for each of the 4 groups of councils - comparing 2012 proforma returns with the Executive's 2003-04 and 2004-05 trend based projections.

17. The main points from the different sets of projections are:-

• For the mainly rural council areas the projected need for permanent accommodation roughly doubles on all projections; from around 3,000 lets per year currently to around 6,000 by 2011-12. Proforma projections are about 700 (12%) higher than the trend based projections.
• In non rural areas (excluding the cities), councils' projections are significantly higher than trend based projections. This difference is mostly driven by differences in the projected number of homeless households discussed above and illustrated in Chart 2b.
• For Glasgow the 04-05 trend based projection is much closer to Glasgow council's own projection and both point to a rough doubling in the number of homeless requiring permanent accommodation by 2012.
• The projections for the remaining cities suggest an increased need for permanent lets from around 3,000 currently to around 5,000 by 2011-12. Notably, most of the projected increase is in Edinburgh.

Chart 7a: Projected need for a permanent let: Rural councils (popn <=50 persons per hectare)

Chart 7b: Projected need for a permanent let: Non-rural areas excluding 4 cities:

Chart 7c: Projected need for a permanent let: Glasgow

Chart 7d: Projected need for a permanent let: 3 cities

Projected need for social lets

18. At present some 4% of the tenancies offered to priority homeless are private tenancies - the remaining 96% being social tenancies. Councils' proforma projections show 8% of the tenancies provided in 2011-12 to be private tenancies. In their report on the 2012 proformas Tribal note that in general private renters do not provide secure tenancies and that therefore provision through a private tenancy may not meet the commitment. The potential for developing the use of private lets has been identified as an issue in the consultation on the commitment.

19. The scope for securing private lets will vary significantly between councils. In assessing the need for social lets the national trend based models simply assume the same proportionate reduction between the need for all lets and social lets as in councils' 2012 returns.

Projected supply of social lets

20. Factors affecting the supply of social lets are trends in the total stock of social housing and the proportion of houses let to new tenants in each year - generally referred to as the re-let rate. For the purposes of the 2012 proforma projections and national modelling the aim is to assess the number of re-lets available from all social housing - council and housing association - in each local authority area. As noted in the Tribal analysis of the 2012 proformas this approach implies that housing association lets will be available to meet the homelessness commitment in the same way as council lets.

Social housing stock

21. Three factors affect the future stock of social housing - demolitions, right to buy sales and new building. As part of Tribal's assessment of the proforma returns, where possible these were compared with returns being made by councils and housing associations on SHQS Standard Delivery Plans and were found to be generally consistent. Thus, the national trend based models assume the same pattern of demolitions and right to buy sales as in councils' proforma returns.

22. The possible level of new social house building in each council area is a major area of uncertainty in assessing the total size of the social housing stock and, as discussed below, the level of lettings available to new tenants. In their proforma returns, councils overall projected around 4,500 new houses for social let in each year up to 2006-07 and around 7,000 per year thereafter. Notably Edinburgh projected an increase of about 1,000 per year in new social house building and Glasgow an increase of about 600 per year from 2007-08. These projections may reflect plans associated with community ownership and related regeneration activity.

23. The national trend based model simply reflects councils' proforma assumptions.

Social lets available

24. The number of social lets available to new tenants in a year is the sum of social lets available from existing stock and the number of new houses built in the year. While a new house might not necessarily be let to a new tenant the letting of the house to an existing tenant will free up a tenancy elsewhere in the landlord's stock.

25. There is a significant amount of uncertainty in projecting re-let rates from existing stock. Factors affecting re-let rates include the desirability of the stock - less popular stock will generally turn over faster than the more popular - and the nature of tenants - young single people will generally spend less time in their first tenancy. Also, lets lost through right to buy sales are not necessarily at the same rate as the re-let rate of the remaining stock, and are probably lower. Nationally during the peak of right to buy sales in the 1980s the number of new lets of local authority houses did not reduce, and the underlying re-let rate increased.

26. In their proforma returns most councils assumed either a constant or slightly increasing re-let rate between 2003-04 and 2011-12. However, some councils have projected significant reductions in re-let rates - reflecting the impact of planned demolitions of low demand high turnover stock. For this analysis the trend based model assumes that re-let rates from the underlying stock remain at the median level of re-let rates in the 3 years between 2001-02 and 2003-04.

27. Charts 8a-d show trends in social lets available for each of the 4 groupings of councils. The main features of the comparisons at this level of grouping of councils are:-

• There is very little overall difference between proforma and trend based projections. The large differences for the few councils projecting a significant reduction in re-let rates do not significantly affect analysis at this level of aggregation.
• For area groupings other than for the 3 cities, at the assumed new building rates the projections show a reduction of around 12% in available lets between 2003-04 and 2011-12.
• The projected slight increase for the 3 cities is driven mainly by the projected increase in new building in Edinburgh.

Chart 8a:

Chart 8b:

Chart 8c:

Chart 8d:

Relationship between projected demand and supply

28. Charts 9a and 9b show, for the rural councils grouping, projected levels of demand for and supply of social housing respectively from proforma returns and from trend based estimates. The key feature of both sets of projections is that the increase in numbers requiring a social tenancy is the main determinant of increases in the proportion of lets required to meet the 2012 commitment.

Chart 9a: Rural councils (50 or fewer people per hectare) Proforma demand/ supply

Chart 9b: Rural councils (50 or fewer people per hectare): Trend based demand/ supply

29. Charts 10a and 10 b compare projected levels of demand and supply in the non-rural areas (excluding the 4 cities). While both sets of projections show increasing proportions of social lets to homeless, the much higher projected growth in homelessness in the proforma returns is the main driver of a much sharper change in the proportion of lets to homeless.

Chart 10a: Non rural councils (excluding 4 cities): Proforma demand/ supply

Chart 10b: Non rural councils (excluding 4 cities): Trend based demand/ supply

30. Charts 11a and 11b compare proforma and trend based projected demand and supply in Glasgow. The projected reduction in homeless applications as a consequence of improved prevention activity is the main determinant of the council's relatively lower projected share of social lets by homeless households compared with the trend based projection.

Chart 11a: Glasgow Proforma demand/ supply

Chart 11b: Glasgow Trend based demand/ supply

40. Charts 12a and 12b compare proforma and trend based projected demand and supply for the remaining 3 cities. Overall both sets of projections show some increase in the proportion of lets to homeless required to meet the commitment. It is important to note that within this grouping the position for Edinburgh is significantly different.

Chart 12a: 3 Cities: Proforma Demand/ supply

Chart 12b: 3 Cities: Trend based Demand/ Supply

Conclusions

41. In comparing the proforma and trend based projections this note has highlighted key areas of uncertainty in projecting changes in demand for social housing arising from the 2012 commitment and in the supply of social lets which will be available in 2012 to meet this. Key points from both sets of projections are:-

• Uncertainty over future trends in levels of homelessness.
• The need for most councils to carefully consider future prevention activity.
• The implications of the 2012 commitment for the mix of household types whose housing needs are to be met - particularly the higher proportion of young single people.
• The need to consider how tenures other than social lets might be used to meet some of these needs.
• The need for councils and the Executive to carefully monitor determinants of and changes in re-let rates of existing stock, including the possible differential impact of loss in re-lets from properties sold under right to buy.

Scottish Executive
Development Department
Analytical Services Division

December 2005