A STRATEGY FOR SCOTTISH SOCIAL SURVEYS
Note of Meeting 5th September 2002, St Andrew's House, Edinburgh
This document is also available in PDF format
Attendees - see Annex
1 Welcome and introductions
Rob Wishart welcomed everyone to the meeting, which had been set up following the ScotStat Board recommendation to consider a more strategic approach to household surveys in Scotland. With a wide range of existing surveys collecting data on different aspects of Scottish households, there is a need to consider whether a more strategic approach could bring benefits to value for money, data quality, timeliness and consistency.
2 Overview of current position and range of possibilities
Bianca Heggie introduced a short paper which outlined the current main surveys covering households in Scotland , and set out the issues concerning possible harmonisation and integration of statistical surveys of households or the household population.
3 ONS Proposal
ONS gave brief outline of their proposal for an Integrated Social Survey. The ONS proposal would be to integrate four ONS run surveys (Labour Force Survey, General Household Survey, Omnibus Survey and Food & Expenditure Survey), covering 120,000 households and costing £11.5 million a year. The Integrated Social Survey would have a core set of questions which would be asked to all respondents, and a selection of modules which would be asked to a selected number of respondents.
The main driver for this work was to obtain the best value for money from these surveys, and from the respondents' time. Other factors such as obtaining statistical coherence on estimates for core indicators and increasing precision of estimates were also important.
ONS were aiming to have the survey underway by 2006 (which coincides with the next review stage for the General Household Survey). They noted that the planning, testing and training of interviewing will cost money, but it was expected that the delivery of results would be quicker. Neighbourhood Statistics was funding a survey (to be launched in autumn 2003), which would be an interim measure consisting of the core questions of the ISS. The Neighbourhood Statistics survey would sample 60000 households.
ONS confirmed that consultation with users would take place before the survey was launched. ONS planned to consult through Steering Group, user groups and user consultations. In future, it was thought that it may be possible to have user group around topic areas (possibly around National Statistics Themes) rather than around surveys.
The development of the ISS requires multi-skilled interviewers. ONS are reviewing the training and contract for their interviewers to ensure that an appropriate workforce is available.
There are ongoing discussions on how to incorporate surveys which are currently run by other departments. Principle issue to address is maintaining the ability to have competitive tendering.
The following points were noted in discussion:
- Care would be needed to keep continuity and time series
- ONS had not had detailed discussion with Northern Ireland yet, but NISRA were aware of the proposal
- Further discussion on the survey coverage in devolved areas was needed
- The idea would be to eventually have the capacity within a local area to cover other issues. More effectively able to meet the needs of the customer.
- The key consideration is the balance between balance/coherence provided by a more integrated social survey , and the flexibility provided by an approach of more specialist surveys.
4 Scottish Social Surveys
The group discussed the issues surrounding better co-ordination of Scottish surveys, including the possibilities of harmonisation and integrating surveys. It was noted that the benefits of the existing framework should be examined further to provide a balanced view. In addition, there would be a need to consider the different aims and objectives of Scottish surveys, and a need to involve policy colleagues (internal and external) to obtain buy in. In particular, there was a need to make sure that individual surveys did not lose the required detail.
It was likely that there would always be a need for on-off specialised surveys. ONS do not view the proposed ISS as an answer for everything, but hope that it will enable them to manage the fieldwork in a more cost-effective way.
The general view was that there has not been enough coherence across surveys since devolution, which has resulted in different estimates being produced from different sources. To try to overcome some of these difficulties, SE has boosted the FRS to get robust estimates on household income, and were currently negotiating a boost for the LFS.
5 Harmonisation
CRU noted that harmonised questions were used already, where they existed and were appropriate. However if a standard (more detailed) set of harmonised questions existed, they could be used for one-off surveys. The need for a core set of economic questions was mentioned.
The following points were made in discussion:
- There was a need for harmonisation on data release and confidentiality
- In England, most modules had harmonised questions already
- Gaining agreements about core questions and the definitions used would be difficult
- The harmonisation of outputs was key for progress
- A need for better quality local information e.g. to support Local Economic Profiles
- A strategic review of surveys must include more than just operational considerations, and should also be driven by information requirements.
It was noted that ONS had dismissed 'adding' the surveys together, due to difficulties with weighting and grossing. However, if co-ordination between surveys could be improved in Scotland, 'adding' the surveys together may be an option.
6 Summary
The group agreed unanimously that there was a general need for better co-ordination between the existing, and any future Scottish surveys (although it was noted that even this carries a cost).
The group agreed with the recommendation to set up a small 'social survey co-ordination committee' which would give further consideration to the issues of co-ordination, harmonisation, integration and the ISS. This committee would be responsible for developing a social survey strategy for Scotland, and begin by working up an initial option for consideration/consultation by the wider group. The committee would be kept to a manageable size (up to 6 members) and consist of representatives from different fields. It would report to, and consult with the wider areas of interest represented by those attending this meeting, and other through the ScotStat register.
The proposed remit for the committee was agreed with the following amendment:
- Single budget should be replaced with co-ordinated funding
SCOTTISH EXECUTIVE
SEPTEMBER 2002
Annex
List of Attendees
Rob Wishart
Pete Whitehouse
Jill Alexander
Duncan Gray
Barbara Doig
Diane Millen
Ruchir Shan
Bob Sargent
Ade Kearns
Mary Sweetland
Ron Mitchell
Sandy Stewart
Pete Knight
Murray McVicar
Ann Millar
David Orr
Peter Craig
Elizabeth Fraser
Robert Wright
Karen Dunnell
Cecilia MacIntyre
Nikki Bennet
Andrew Fleming
Hervey Gibson
Katherine Hudson
Tony O'Sullivan
Bianca Heggie
Jenny Boag
Susan Provan
Apologies
Alan Campbell
Peter Scrimgeour
Gillian Raab
Andrew Goudie