2012 homelessness target marion gibbs and duncan gray modelling seminars - edinburgh
TRANSCRIPT
2012 Homelessness Target
Marion Gibbs and Duncan Gray
Modelling seminars - Edinburgh
Background
• 2012 target
• 2009 interim target
•Reached or exceeded by 14 LAs
•A further 5 LAs reached or exceeded it during one or more quarters
•Priority assessments in 6 LAs were 10 or more % points below 2009 target
2012 SG/COSLA Steering Group
• Joint Steering Group started meeting in October 2009
• Membership – CoSLA (chaired by Cllr Brian Goodall, Cllr Harry McGuigan and officials), SG (Minister and officials), ALACHO, SOLACE and SFHA
Remit
• To assess, inform and influence progress towards the 2012 homelessness target. To oversee ongoing and planned work to assess progress against the interim targets set for local authorities for 2008-09 and to determine the implications for further action needed to meet the 2012 target
Workplan
• Four main areas agreed:
•Continued leadership at both political and corporate level – promoting joint working
•Preventing homelessness
•Ensuring access to existing stock among PRS and RSLs
•Investing in appropriate areas
Increasing homelessness applications
Applications for assistance
0
10,000
20,000
30,000
40,000
50,000
60,000
70,000
Increasing priority falling non-priority
Priority and non-priority homelessness
0
5,000
10,000
15,000
20,000
25,000
30,000
35,000
40,000
45,000
50,000
1992-93
1993-94
1994-95
1995-96
1996-97
1997-98
1998-99
1999-00
2000-01
2001-02
2002-03
2003-04
2004-05
2005-06
2006-07
2007-08
2008-09
Priority Non-priority
Increasing numbers getting permanent accommodation
Accommodation secured by homeless applicants
0
2,000
4,000
6,000
8,000
10,000
12,000
14,000
16,000
18,000
20,000
Social Rented Return to previous/ present accommodation Hostel etc Other
Analysis based on the year the application was completed
Analysis based on the year the household applied for assistance
Homeless are mainly young
Homeless by age of main applicant: 2008-09
0
200
400
600
800
1000
1200
1400
1600
1800
2000
Relationship breakdown is the main driver of homelessness
Main groups of homeless by prior circumstances and reasons for homelessness
Temporary accommodationSupported
accommodation
Prison, hospital armed forces
Rented/ Owned: Financial/landlord
reasons
Rented/ Owned: Dispute etc.
Friends/ partners: Dispute
Other Parents/ relatives: Dispute/ Asked to
leave
required
spend
total
budget
Extra budget
required
Progress against interim targets
Progress towards 2012 target by Local Authority
-10 0 10 20 30 40 50 60 70 80 90 100
Aberdeen City
Aberdeenshire
Angus
Argyll & Bute
Clackmannanshire
Dumfries & Galloway
Dundee City
East Ayrshire
East Dunbartonshire
East Lothian
East Renfrewshire
Edinburgh
Eilean Siar
Falkirk
Fife
Glasgow City
Highland
Inverclyde
Midlothian
Moray
North Ayrshire
North Lanarkshire
Orkney
Perth & Kinross
Renfrewshire
Scottish Borders
Shetland
South Ayrshire
South Lanarkshire
Stirling
West Dunbartonshire
West Lothian
% of homeless assess as priority
'03-04 start position Position reached by08_09
Issues looked at to date
• 14,847 applications assessed as homeless from <25 (37% of all assessed as homeless)
• 8 LAs have more young women assessed as homeless than young men (Aberdeen, Dundee, East Dunbartonshire, East Lothian, Edinburgh, Eilean Siar, Perth and Kinross and West Dunbartonshire)
Key Statistics
• 22% of <25s are single parents (24% for homeless population as a whole)
• 5% of homeless <25s are couples with children
• 2% had leaving supported accommodation as last form of accommodation
Areas where high percentage of young people
• Orkney (60%)
• Moray (48%)
• West Lothian (44%)
• Fife (43%)
• Clackmannanshire (42%)
• Angus and North Ayrshire (41%)
Areas where low percentage of young people
• Glasgow (27%)
• Inverclyde (28%)
• East Renfrewshire (32%)
• Eilean Siar (32%)
• North Lanarkshire (33%)
Dumfries and Galloway
• 40% of population under 25
• 2.8% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (69%); single parent (19%); couple (6%); couple with children (6%)
• Dispute non-violent or asked to leave – 57.5%
• Young people in area – 15-24 – 10.8%
• SHR report – C – 2006/07
• Youth unemployment – 33.2%
East Lothian
• 37% of homeless population under 25
• 2.9% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (67%); single parent (17%); couple (9%); couple with children (7%)
• Dispute non-violent or asked to leave – 68.3%
• Young people in area – 15-24 – 11.9%
• SHR report – C – reinspection 2007/08
• Youth unemployment – 30.6%
Edinburgh
• 35% of homeless population under 25
• 2.3% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (67%); single parent (23%); couple (6%); couple with children (4%)
• Dispute non-violent or asked to leave – 66.2%
• Young people in area – 15-24 – 15.3%
• SHR report – A – 2005/06
• Youth unemployment – 27.0%
Falkirk
• 39% of homeless population under 25
• 4.7% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (62%); single parent (18%); couple (13%); couple with children (6%); other (2%)
• Dispute non-violent or asked to leave – 53.7%
• Young people in area – 15-24 – 12.1%
• SHR report – D – 2007/08
• Youth unemployment – 29.9%
Midlothian
• 37% of homeless population under 25
• 2.7% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (62%); single parent (20%); couple (9%); couple with children (7%); other (2%)
• Dispute non-violent or asked to leave – 65.7%
• Young people in area – 15-24 – 12.7%
• SHR report – D – 2007/08
• Youth unemployment – 35.1%
Orkney
• 60% of homeless population under 25
• 1.7% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (85%); single parent (3%), couple (6%); couple with children (6%)
• Dispute non-violent or asked to leave – 60.6%
• Young people in area – 15-24 – 19%
• SHR report – D – 2004/05
• Youth unemployment – 30.0% (Scotland 29.4%) for 18-24 – JSA claimants
Scottish Borders
• 37% of homeless population under 25
• 2.7% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (67%); single parent (13%); couple (13%); couple with children (7%)
• Dispute non-violent or asked to leave – 56.4%
• Young people in area – 15-24 – 10.6%
• SHR report – C – 2007/08
• Youth unemployment – 30.9%
Shetland
• 40% of homeless population under 25
• 2.7% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (74%); single parent (11%); couple (8%); couple with children (6%)
• Dispute non-violent or asked to leave – 62.9%
• Young people in area – 15-24 – 11.4%
• SHR report – C – 2007/08
• Youth unemployment – 23.7%
Stirling
• 38% of homeless population under 25
• 1.8% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (75%); single parent (16%); couple (5%); couple with children (4%)
• Dispute non-violent or asked to leave – 70.6%
• Young people in area – 15-24 – 15.3%
• SHR report – C – 2003/04
• Youth unemployment – 29.6%
West Lothian
• 44% of homeless population under 25
• 3.0% of 16 – 24 population = assessed as homeless (Scotland 2.3%)
• Young homeless: single person (66%); single parent (18%); couple (8%); couple with children (7%); other (1%)
• Dispute non-violent or asked to leave – 66.5%
• Young people in area – 15-24 – 15%
• SHR report – C – 2005/06
• Youth unemployment – 31.7
Lets to homeless households - LAs
• 45% of local authority lets in Scotland going to homeless applicants:
•East Lothian Council - 53% (missed target by 7%)
•Edinburgh – 60% (missed target by 2%)
•Falkirk – 60% (met target by 1%)
•Midlothian – 44% (missed target by 5)
•Orkney – 37% (met target)
•Shetland - 30% (missed target by 11%)
•Stirling – 55% (missed target by12%)
•West Lothian – 70% (missed target by 6%)
Lets to homeless households - RSLs• Across Scotland – 22% of RSL lets to
homeless households (s5 and homeless nominations) – APSR figs (25.5% SCORE)
• Variation in this:
•55% of lets to under 5% of lets
•Full stock transfers also vary – 50% in DGHP to 24% for both SBHA and River Clyde Homes
•GHA – 29%
Big increase in households in temporary accommodation
Homeless households in temporary accommodation: 31 March
0
2,000
4,000
6,000
8,000
10,000
12,000
1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009
All households Households with children
Question
• What are the drivers of change in homelessness levels and what are the barriers to achieving the 2012 target?
2012 Modelling (Waugh Model)
(Simplified Spreadsheet Model)
Assessing councils’ capacity to meet 2012 homelessness
commitment
Aim of presentation
• To give a broad overview of model and set out key features/ assumptions.
• To identify issues arising from use of model over past couple of years.
• To discuss development/ future use/ variants.
Context
• 2012 commitment.
• Assist Ministers in their statutory duty to assess capacity of each council to meet the commitment.
• Assist Ministers in working jointly with councils to assist all to ‘get into a better position’ to achieve 2012.
What the model does(a) Projects supply of lets to meet the needs of
priority homeless for each year to 2015-16. [2012 and beyond.]
(b) Projects demand for lets for homeless under a range of assumptions.
(c) Projects number of LA/ LSVT; RSL; PRS lets taken by homeless in each year under a range of assumptions.
(d) Projects the number in temporary accommodation and the amount of time spent in temporary accommodation in each year.
Outline of the model
LA/ LSVT Stock New Build
Sales
Demolitions
SOCIAL STOCK Social Lets
Relets of existing stock
+
==Available LA/
LSVT Lets
-
Decants
New BuildSales
Demolitions Relets of existing stock
+
==Available RSL
Lets
-
Decants
HOMELESSNESS
Assessed Priority Assessed Non-priority
Social lets available/
needed Other destinations/ outcomes… e.g. Returned to Previous Accommodation. Moved in with friends/ relatives. Made own arrangements. Lost contact.
Private let.
RSL Stock
Note:- The red boxes show the main outputs from the model
+ =Temporary
AccommodationNeeded
Temporary accommodation
as outcomeTemporary
accommodation awaiting let
Key inputs: Social lets
• Projected supply of social lets comes from turnover of existing stock, new building demolitions including decants.
• Separate projections for – LA/ LSVT – RSL
• Allows modelling of impact of moving to equal shares of lets to homeless.
Key inputs: Private Rented Lets
• Very simple set of assumptions:-– Estimated turnover of PRS from SHS and
PRS registration sources.– Assume that PRS let would be suitable for no
more than x% of homeless. [Currently 20%, can be
varied.]– Assume that no more than y% of PRS lets
would be suitable for homeless. [Currently 10%,
can be varied.]
Key inputs: Homelessness
• Numbers homeless and proportion in priority in base year (now 08-09).
• Shape of profile to achieve 2012 [gradual v big-bang.]
• Impact of prevention over projection period.• Impact of drop-outs:-
– Maximum % of priority homeless who will need a permanent let.
– Reduction in drop-out rates over projection period.
What an output workbook looks like
Advantages
• Detailed profile of stock, lets and homeless levels.
• Identifies and incorporates all the key factors affecting balance between need and supply.
• Sophisticated mathematical model providing a projection of volume of temporary accommodation needed.
Limitations• Takes about 5 hours to run the model for all
councils.• The Scottish Government Version doesn’t allow
single council runs.• Can’t readily vary the profile of % homeless
assessed as priority.• Can’t put restrictions on % of LA/ LSVT/ RSL
lets to homeless.• Uses MATLAB so can’t be provided to councils.
Simplified spreadsheet based model
• Uses almost all the same inputs, taken directly from Waugh model inputs.
• One model for each council.• Projects balance between need and supply
under the given set of assumptions in each year to 2015.
• Allows restrictions on % of lets to homeless.• Doesn’t project numbers in temporary
accommodation.
Example input sheet
Example output summary
Points for discussion
• What factors (for your LA) might affect the use of standardised assumptions in the models?– Number of LA/ LSVT lets available.
• Reprovisioning/ decants.
– Number of PRS lets available.– Homeless prevention.– % of priority homeless requiring a let.
• Value of projected use of temporary accommodation.
Use of 2012 modelling for Strategic Housing Investment Working
Group
Presentation will cover
• Purpose of the SHIF Working Group.
• Likely use of 2012 modelling.
• Issues arising.
SHIF Working Group
• Joint CoSLA Scottish Government.• Remit is to provide advice to Ministers on
criteria to use to distribute capital grants (mainly development programme).
• The main drivers of affordable housing need are:-
• Addressing wider affordability;• Supporting regeneration;• Meeting 2012 homelessness commitment.
SHIF Working Group
• Working towards a distribution formula based on indicators relevant to each driver of need.
• No final decisions on either the indicators to be used or the weights to be applied to these.
Use of 2012 models• Runs of Waugh and related models show – under the specific assumptions
used - relative investment needed to ensure that no more than X% of social lets will be needed in 2013-14 for homeless
– 2013-14 is first full year after Dec 2012.– X% has generally been set at 60%.– Homeless has been based either on latest year or on a given % per year
reduction due to prevention.– Projected stock and lets has been on set assumptions about:-
• Turnover of existing LA/ RSL normal lettings stock.• RTB sales.• Decants from non-viable stock.
– Proportion of priority homeless not requiring a social let.– Standard assumptions about potential for use of PRS.
• Councils with largest ‘shortfall’ in lets from Waugh model generally [but not exactly] are also councils with net affordable need in wider affordability assessments (e.g. Bramley).
Issues (1)
• Wide variations between councils in patterns and incidence of homelessness which can’t be easily explained by external factors.– Regression analysis shows that relative levels
of deprivation and constraints on affordable supply do play a part.
– But a large amount of unexplained variation.• See next 2 slides.
Higher levels of income deprivation imply higher levels of homelessness: But significant
unexplained variationHomeless as % of all households v % of total population income deprived
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
0 5 10 15 20 25 30
% of popn income deprived
Ho
mele
ss h
ou
seh
old
s a
s %
of
all
ho
useh
old
s
Adding a factor for wider affordable need improves the fit but only slightly
Regression % homeless v multiple regrssion on inc deprived and Can't afford as % of lets
0.00
0.50
1.00
1.50
2.00
2.50
3.00
3.50
0.0 0.5 1.0 1.5 2.0 2.5 3.0
Predicted from regression model
Actu
al
ho
mele
ss a
s %
of
all
ho
useh
old
s
Issues (2)
• Homelessness varies significantly from year to year by council area.
• Between 2005-06; when homelessness peaked in Scotland; and 2008-09 homelessness decreased by over 20% in 8 council areas and increased by over 20% 8 council areas.– See next slide.
Also homeless incidence varies a lot over time
Percentage change in number of households assessed as homeless between 2005-06 and 2008-09
-80 -60 -40 -20 0 20 40 60 80
Aberdeen City
RenfrewshireShetland
Scottish BordersSouth Lanarkshire
Dundee CityFalkirk
Stirling
East LothianAberdeenshire
MidlothianEilean Siar
EdinburghWest Dunbartonshire
East DunbartonshireSouth Ayrshire
HighlandAngus
East RenfrewshireDumfries & Galloway
MorayNorth Ayrshire
East AyrshireNorth Lanarkshire
Argyll & Bute
West LothianFife
ClackmannanshireGlasgow City
Perth & KinrossInverclyde
Orkney
Progress of SHIF discussions
• Likely to recommend not using current homelessness levels, but rather:-– using projected homeless levels from a base
at around 2005-06; and – Projected year on year reduction to reflect
councils’ capacity to reduce homelessness through effective prevention.
Further SHIF related work
• Credibility assessment alongside other indicators.
• Reviewing/ checking some of the standardised assumptions on supply.
Points for discussion (1)• How do we ensure that approach doesn’t
penalise effective prevention?• What reasons might there be behind big
year-on-year changes in homelessness levels? Changes in:-– Underlying drivers of homelessness.– Applicant behaviours.– Council behaviours/ policies.
Points for discussion (2)
• Views on capacity of PRS.– How can we improve our modelling on this?
• Constraints on % social lets to homeless?
• Any other issues?
Session 4
• 2012 Steering Group – discussion around measuring prevention activities
• But also more than this – measuring the impact of prevention
• Some local authorities are definitely focussing on prevention, but homelessness increasing
Session 4
• How best can we effectively record prevention activities?
• How can we measure the impact of prevention?
• How can we monitor success?