The Scottish Government - PowerPoint PPT Presentation

the scottish government n.
Skip this Video
Loading SlideShow in 5 Seconds..
The Scottish Government PowerPoint Presentation
The Scottish Government

play fullscreen
1 / 18
Download Presentation

The Scottish Government

Presentation Transcript

  1. The Scottish Government Using Administrative Data in the Housing Supply Indicator Claire Boag, Communities Analytical Services

  2. Outline • Who we are • Pros and cons of administrative data • NI 32 – Increase the rate of new house building. What does that actually mean? • New build components • Conversions • Refurbishments • Things to consider

  3. Housing Statistics branch • Collect, analyse and disseminate housing data • 20 administrative returns from local authorities: • Local authority housing management • Changes to stock levels • Right to buy sales • New house building • Also have access to administrative data collected by others: • Housing Benefit / Market Evidence • Housing Association new build • Landlord Registration

  4. Advantages of administrative data • Low costs of obtaining data from an existing source. • Faster and more frequent analysis may be possible. • In particular, the time between an event occurring and being recorded • Eg - most housing stats data available immediately (exceptions include private new build) • frequency – eg. quarterly figures for local authorities could be difficult to achieve with a survey without huge expense • Sometimes more detailed figures are available • Eg - for geographical areas where data has geographic referencing • No statistical sampling errors or non-response bias. • Doesn’t rely on memory of respondent.

  5. Disadvantages of administrative data • Effort may be needed to make administrative sources useable for statistical purposes. Eg – estimating missing values, getting rid of redundant data, linking to other data sources • Lack of independence – not controlled by statisticians. • Definitions and classifications may often not be ideal for statistical purposes. Eg - data on housing association starts is actually based on approvals. • Coverage of the data may not be complete and it may be difficult to make comparisons with other statistics. Eg – landlord registration database Eg – HB claimants from RRS v DWP • Changes in provider data systems may cause discontinuities. Eg – housing list cleaning or system improvements • There may be some danger of interruption to the supply of data. Eg - staff turnover in LAs Eg – IT problems • There may be constraints on the use of data Eg – Market Evidence Database

  6. Housing supply indicator • Data on housing supply informs National Indicator 32 (Increase the rate of new house building), and comprises the following elements: • new house building: houses completed by or for housing associations, local authorities or private developers for below market rent or low cost home ownership; houses completed for market sale by private developers. • refurbishment: houses acquired by housing associations and refurbished either for rent or low cost home ownership. Refurbishment of private dwellings funded wholly or partly through the Affordable Housing Investment Programme. • conversion: new dwellings created by conversion from non-housing to housing use.

  7. Housing supply indicator Terminology issues • Increase the rate of new house building • Uses absolute numbers not rates • Includes new housing provided through other means

  8. Private new build data • Data provided on a quarterly basis by local authorities via the NB2 return. • Can be based on different data/methods in each LA though each should be consistent over time. • No marked seasonality but data for individual quarters is volatile therefore quarterly change unreliable

  9. Local authority new build data • Data provided on a quarterly basis by local authorities via the NB1 return. • Council house building limited in recent years, therefore numbers small and data volatile • Errors fairly unlikely since numbers low

  10. Housing association new build data • Data provided on a quarterly basis by HID colleagues from the AHIP database • Data provided by those involved with managing the developments therefore quality higher than private new build. • Very volatile

  11. Annual new build • Annual new build data much less volatile – change more clearly shown • Issues - Need to watch out for duplication across sectors • LA new build with AHIP funding • Purchase of private new build by LA for social rent

  12. Conversions • Collected from local authorities through annual return

  13. Refurbishments • This is proxy data – we only have information on housing association refurbishments and private refurbishments receiving public subsidy. • All other private refurbishments (resulting in addition to housing supply) are excluded

  14. Things to consider (1) How and why was the data collected? • Private and local authority new build • Paper based returns, produced purely for the SG. Data tends not to be heavily used within the councils (with a few exceptions). • Labour intensive, burdensome return • Based on a combination of completion certificates, site visits and council tax assessors data • Smaller, more urban local authorities are more likely to use site visits, large rural authorities need to rely on completion certificates • There is anecdotal evidence that there are issues with completion certificates in some councils, eg: self-builders have no incentive to apply for a certificate. Often long delays between ‘effective’ completion and legal completion.

  15. Things to consider (2) How and why was the data collected? • Housing association new build and refurbishments • Information collected by SG Housing Investment Division to monitor Affordable Housing Investment Programme. • Data extracted from live database, and HID carry out quality assurance checks before passing it on. • Delays in the recording of completions mean that the data is subject to annual revisions each March, therefore quarterly or calendar year monitoring inappropriate. • Conversions • Information on conversions are provided by local authorities on the annual Stock4 return. • Data is provided in summary form, therefore there is little scope for us to check the accuracy, other than to compare with other LAs and over time • Difficulty collecting information from local authorities

  16. Things to consider (3) Is there a consistent time series of estimates? • Private and local authority new build • Information has been collected in its current form since 1986. • Issues with data for Highland – new source used, but revisions made back to 2000 so no issues with consistency in recent years. • If a local authority changes the way it collects or records the information, we will only know if they tell us. • Housing association new build and refurbishments • provided since 1992 • Changes afoot with HA starts – we will need to consider creating consistent back series • Conversions • Consistent at least since 2001. Prior to that was collected as part of new build data. No obvious inconsistencies

  17. Summary • Know how the data is collected • More scope for error with paper based collection • Data that is difficult to collect will be more error prone • Know why the data is collected and how it’s used by the provider • The more it’s used by the data provider, the more accurate it should be • Understand the nature of volatility in the data • Is it a real change or a data artefact? • Investigate unexpected peaks and troughs • Know whether there are definitional or system changes that would cause a discontinuity in a time series

  18. Contacts Email, or Telephone: 0131 244 7234 Fax: 0131 244 0446 Post: Communities Analytical Services (Housing Statistics) Area 1-F Dockside Scottish Government Victoria Quay Edinburgh EH6 6QQ