Homelessness Statistics User Group meeting paper: March 2023

HL1 update – Ukrainian displaced people

HL1 was updated in September 2022 to better capture Ukrainian displaced people (UDP) seeking homelessness support

Response options were added for the following:

• eligibility – They are Ukrainian nationals, lawfully present in the UK through the Ukraine Family Scheme or Homes for Ukraine (private or super sponsor) visa scheme (and had been living in Ukraine on or before 1 January 2022)
• ethnicity – Ukrainian [also updated to align with 2022 Census]
• postcode of last settled address – UKR

As at March 2023:

• 24 local authorities have either confirmed their systems are updated or have submitted data returning one or several of the new response options
• no data was published as part of the April to September 2022 statistics update [on 31 January 2023] due to incomplete national coverage
• aim to have near complete coverage by receipt of quarter 4 data [January to March 2023] backdated for previous cases as relevant

Are there any blockages to this that we are not aware of?

• know there are some local authorities (LAs) are waiting for software providers to resolve and others have larger systems update in the pipeline which will include new response options

Data quality:

• we monitor the returns and quality assure data quarterly [manually at the moment]
• instances with case submitted using one of the new response options but not another – when queried so far has been updated

Are there any difficulties or barriers to recording this data?

Unsuitable accommodation and breaches of the unsuitable accommodation order (UAO)

Unsuitable Accommodation Order: A breach should be recorded when a household has been in unsuitable temporary accommodation for more than 7 days

In Spring 2020, it became apparent there was uncertainty and inconsistency in the reporting of breaches due to:

• change in legislation (from households with pregnant women/children only to all)
• COVID-19 exceptions

Scrutiny of the data unveiled:

• errors in reporting
• inconsistency in reporting

Errors and inconsistencies

• some reported as a breach, when not marked as unsuitable
• certain types of accommodation not being marked as unsuitable, when it was expected it should, and vice versa
• breaches for durations less than 8 days
• no breaches for durations more than 8 days
• not the same for all – some placement level, some application level
• difficult to determine for some

Action to resolve

• discussion with policy and working groups
• consulted legal colleagues and Scottish Housing Regulator
• decision was reached that reporting should be consistent with wording of legislation i.e. application level
• update data collection guidance, including worked examples    
  • in time for data returns from 1 April 2022

Current data

• some LAs have implemented changes and are reporting in line with enhanced guidance
• however, not all LAs have been able to do this currently
• manual checks of data each quarter
• follow up with local authorities
• implementing automated checks
• not all errors identified have been rectified
• ongoing process, over the transition period

Additional HL3 questions 12 and 13

In 2019, two additional questions were introduced to the HL3

• these related to whether the temporary accommodation placement included a pregnant women or children
• a majority of LAs have implemented these additional questions, but not all
• we are looking to include these figures in our upcoming annual 2022/23 publication – we want to have a full set of data from 1 April 2022 from all local authorities

Progress

• 26 out of 32 LAs have provided this data – well done!
• all LAs who completed question 12 completed question 13 correctly
• four LAs require backdating
•  as there are cases belonging to quarter 1 and quarter 2 which are still incomplete
• two LAs have still been unable to provide this data 
• both of which have confirmed they will be in a position to do so in quarter 4
• there will be backdating efforts for the upcoming quarter 4 collection where we will be in touch with relevant councils to assist

HL2 vs HL3

HL2: temporary accommodation snapshot at the end of the quarter 
HL3: temporary accommodation placements
Should be able to recreate HL2 figures from HL3

Differences exist between HL2 and HL3

• nationally, HL3 is 2%-4% higher than HL2
• although, it’s not always the case that the HL3 returns are higher
  • ~2/3 LAs HL3 higher than HL2, ranging from 2% to 34%
  • ~1/3 LAs HL3 lower than HL2, ranging from 2% to 8%

To note, 9 LAs HL2 and HL3 figures are extremely close

Reasons for differences

• data not collated the same way, or from the same source i.e. the HL2 return is not generated from the placement level data
  • what can be done to bring these figures more in line?
• not all accommodation types are included in each return
  • why aren’t all accommodation types included in both sets of returns?
• missing exit dates for HL3 placements, therefore these placements appear to be ‘live’ at the end of quarter, artificially inflating the live HL3 figures
  • what is preventing exit dates being provided?

Next steps

Targeted discussion with LAs to understand more about the nature of their differences to determine if:

(a) they can theoretically match or (b) they will never be able to theoretically match

If a) then working with LAs to get these to match

If b) data review work to pick up

Temporary accommodation (TA) average duration

Many LAs have asked how Scottish Government calculates:

•  ‘duration’ (placement level)
•  ‘total duration’ (household level)
•  ‘average total duration’ (local authority and national level)

This table [see supporting document] is from our annual publication [image of table 36]

• it shows the average duration (days) that households have spent in TA for each LA by financial year of case closure
• next slides will go through step by step how these are calculated

Calculating duration

Placement level duration: duration = date of exit - date of entry
e.g. 12.01.2023 - 10.01.2023 = 2 days
Household level duration: total duration = sum of all placement durations belonging to a single household

• only for placements where offers have been accepted and taken up
• if date of exit is missing, but the homelessness application has been closed, the close date will be used instead
• a household's placement history can span back over years

The average duration will be calculated from the total durations of all households which have closed within the financial year
[see supporting document for graphical representation of how to calculate average duration from multiple household durations that are all made up of one or several placement durations]

Placements with zero day duration

There are placements which have ‘0 day’ duration (entry date = exit date)

• 5 LAs have a reasonable number of these records (>50 over last 3 years) 
• 17 LAs with a smaller number of these records (<50 over last 3 years)
• this impacts all of our published HL3 temporary accommodation outputs

Are these recording errors or are these genuine placements?
Why would a placement be 0 days?
We will be reaching out to LAs to query these ‘0 day’ duration placements
This will help us understand if a data processing solution is available

Postcode and national insurance number

• quality of these discussed at the last HSUG meeting (regarding completeness)
• not progressed as other areas have been prioritised
• will be picked up as wider data review work
• consideration of use in relation to data linkage
• further geographical analysis, e.g. linking to Scottish index of multiple peprivation (SIMD)

[Flash feedback from data providers was sought on the use of quarterly reports and data outputs]