Minutes of the 14th Meeting of the Output, Geography and Confidentiality Working Group (OWG) Held on Friday 5 March 2004 at Skipton House, London.
Members Present:
Frank Thomas (Chair)
Simon Hazlewood (Sec.)
Brian Green
Chris Denham
Pete Large
Ian White
Brendan Davis
Lucy Baker
Alexa Courtney
David Greeno
David Martin
Eileen Howes
Jenny Boag
Keith Dugmore
Nick Walker
Oliver Duke-Williams
Paul Stoker
Richard Price
Roger Morgan
Stephen Reynolds
General Register Office for Scotland
General Register Office for Scotland
Northern Ireland Statistics & Research Agency
Office for National Statistics
Office for National Statistics
Office for National Statistics
Office for National Statistics
Office for National Statistics
Office for National Statistics
Department of Health
University of Southampton
Greater London Authority
Falkirk Council
Demographic Decisions Ltd
UK Online
University of Leeds
Department for Education and Skills
Experian Business Strategies
Royal Burgh of Kensington & Chelsea Authority
Department for Transport
Apologies were received from:
Ann Hockey
Justin Hayes
Colchester Borough Council
University of Manchester
Papers: OWG(04)01 - 2001 Census Debrief and 2011 Census (page 11) OWG(04)02 - The 2011 Census: user's comments on the proposed design (page 13)
Action Points: Action points arising from the meeting are listed on page 10.
1.
Chairman's introduction
1.1
Frank Thomas welcomed everyone to the 14th meeting of the OWG and noted apologies received from Justin Hayes and Ann Hockey.
2.
Minutes of last meeting
2.1
The minutes of the last meeting were passed without amendment.
3.
Progress made/work outstanding
3.1
Workplace statistics
3.1.1
Frank Thomas confirmed that a decision had been reached in all 3 Census Offices with regard to the workplace tables. In Scotland, the decision had been to release the tables (UV75-UV79) at Output Area level, applying small cell adjustment (SCA) to the figures in the tables. This method had not been applied to any other Scottish tables. He then said that the tables had now been released on SCROL (Scottish Census Results OnLine) and that they would be released on SCROL CD7 (travel, migration and living rent free tables) in the next month or so.
3.1.2
Chris Denham said that a review of potential disclosure risks on workplace tables had commenced in ONS last October. It was also confirmed that delays had been experienced in producing tables on migrants due to technical problems with SuperCROSS tables. This meant that remaining releases of Census results for England and Wales, most of which were expected to contain workplace and out migrant tables, had not gone ahead as scheduled.
3.1.3
ONS had hoped to make an announcement prior to the OWG meeting but confirmed that the Industry and Occupation workplace tables would not be released for Output Areas. The remaining tables would be released as planned and the position would be outlined in the next Census News. This would include the release of CAS tables in April.
3.1.4
Brian Green said that NISRA had released all their workplace tables (down to ward level), together with migration and travel tables on 17 February. He highlighted that an expanded data set had been provided in recognition of user demands and interest in such data and pointed out that the migration tables had been specifically tailored to provide additional detail at the various levels of geography. Brian continued that the downside of this was that a workaround was now required to load the tables into NICA (www.nicensus2001.gov.uk) which was developed on the basis that a common table template would pertain across all levels of geography.
3.1.5
Frank Thomas said that such differences that had emerged between the Census Offices were largely due to different interpretations of the application of rules covering business statistics to the census workplace tables.
3.1.6
On the technical problems for SuperCROSS in the development of these and the other travel and migration tables, Frank Thomas explained that the process required all 3 Census Offices to populate various parts of tables on each other's behalf applying SCA or not as appropriate. These parts had then to be joined with parts populated on their own databases and made ready for release. This part of the process had worked well and had shown good communication between all involved.
3.1.7
David Martin asked where Super Output Area (SOA) statistics fit in with workplace tables, Chris Denham said that the SOA statistics would be covered by the same rulings on confidentiality as applied to the workplace tables.
3.2
Origin Destination
3.2.1
Frank Thomas said that delays had been experienced with the release of origin destination data. If a single release were to go ahead this would now be in July. The Census Offices were now considering a phased release. He confirmed that the files would be provided in csv format and that this would be run by ONS so no (Scottish) csv formatting issue should arise.
3.2.2
Eileen asked if details of the revised tables could be sent to her colleague Hywel Davies. Frank Thomas replied that these would be available via test data produced by Alan Fleming (GROS) and that he would arrange for her to receive a copy. Richard Price and Oliver Duke-Williams also requested copies.
3.2.3
Oliver Duke-Williams asked when the Special Travel Statistics part of origin destination would be released. Frank Thomas replied that these would be released at the same time as the Special Workplace Statistics.
3.2.4
Oliver Duke-Williams then asked whether the matrices could cut across scales, such as Ward to Output Area. Frank Thomas confirmed that this point had arisen during internal discussion and 'intermediate files' would be retained for such output to be produced after initial release (IT resources permitting).
3.3
Area Statistics
3.3.1
Pete Large said that ONS had released the 2nd part of the National Report (Local Authorities) in February. It was planned to release Civil Parish and postcode sector data in March with urban areas planned for May.
3.3.2
Chris Denham confirmed that because of the popularity of Parish data, especially in Wales, it would be put on Neighbourhood Statistics (NeSS) also.
3.3.3
He also said that ONS had not received a lot of feedback following the release of the National Report and that member's comments would be welcomed.
3.3.4
Pete Large continued by confirming that both postcode to Output Area and Output Area to Higher Areas files were now available from ONS. He then said that the supporting information (metadata) had been made available on the ONS website as each section had been completed. A paper publication bringing all the supporting information together, Definitions Volume, would be released in the summer.
3.3.5
Pete Large also confirmed that ONS CAS migration tables and KS24 would be released in April for Output Areas.
3.3.6
In addition, once all the pre-announced statistics had been released, ONS would undertake a rationalisation exercise to compile any missing data or errors and release this on a CD or DVD product. It was also confirmed that ONS had made considerable progress with the evaluation reports from the 2001 Census.
3.3.7
Nick Walker and Richard Price both raised questions on the treatment of small postcode sectors in ONS output. Pete Large explained that such small sectors would be subsumed in neighbouring areas, and offered to supply more information after the meeting if required.
3.3.8
Brian Green said that NISRA were currently working on the release of their Theme tables. He highlighted that a total of 25 tables would be released and that the NI level version of these had been passed to their publishers for inclusion in a NI level report. He added that while the expected release date was late March, an official announcement had not yet been made. In terms of output on CD ROM, Brian summarised that to date NISRA had released their KS, ST and CAS CD products. He was hopeful that both the Univariate and Migration, Travel to Work and Workplace Population CD products would be available for release late March/early April, with the Theme tables CD following in April/May.
3.3.9
He also confirmed that consultation on the Grid Square product for Urban Rural statistics was underway and that they were close to a definition of Urban/Rural areas. Once this had been completed they would look at the data to be released but that this would not be released before the remaining Area Statistics.
3.3.10
Frank Thomas then explained the position for GROS, confirming that following the release of all travel, migration and living rent free tables on SCROL, all Area Statistics for Scotland were now available.
3.3.11
The travel, migration and living rent free tables would also constitute SCROL CD7, a master of which was currently being put together by Geowise. This would hopefully be released in early April.
3.3.12
He also confirmed that GROS would also re-supply SCROL CDs 1-6 to replace any data errors arising from the 1st release in 2003. This was being done in-house by GROS and would follow release of SCROL CD7. These products would then be pulled together in a DVD product which could then be seen as the 'final version of Area Statistics'.
3.3.13
Frank Thomas said that work was also underway to complete the SCROL Data Warehouse and Bulk Delivery service by adding the travel, migration and living rent free tables. Both these products would be in the same csv format as before.
3.3.15
GROS was a little behind ONS in terms of 2001 Census evaluation, but that this would be looked at soon and would be made available on the GROS website over the next few months.
3.3.16
Brian Green confirmed that the NICA site, which is largely based on SCROL, has been publicly available from November of last year and that work was currently underway to incorporate the Thematic Mapping module. He added that completion of this and work on the output media front would allow the work on the 2001 Census evaluations to move further up the agenda.
3.3.17
David Greeno asked when the Standard Tables for Health Boards for England & Wales would be made available. Chris Denham said that this would be March but that Pete Large would let David know the exact position.
3.4
Other Products
3.4.1
Pete Large confirmed that the first 'Focus On...' Reports had been published on the ONS website in January and that work was ongoing with other reports. The multi-source topic report tables, that feed data into the 'Focus on' reports, are available free of charge from customer services. David Greeno asked when the other reports were due and Pete said he would check the position.
3.4.2
Eileen Howes said that Hywel Davies was converting these tables into SASPAC system files, but that they had to add in their own footnotes to help users understand the data. Eileen Howes also requested that the GLA be put on a 'standing order' for these tables as she had experienced delay in receiving the data. Pete Large said that there was a list of these tables on the ONS website that was updated twice monthly, and that he would organise something with Eileen.
3.4.3
Chris Denham said that ONS had also published the Local Authority level classification and that the ward classification was due out at the end of March though there would be the usual problems in devising simple descriptions of each group in the classification. This would apply also to the Output Area level classification. Chris added that the Local Authority level classifications show both striking differences between 1991 and 2001 and polarisation in lowland England, particularly in London.
3.4.4
Keith Dugmore said he was glad that these would be made freely available and welcomed the use of SVG (scalable vector graphics) to present the data.
3.4.5
Chris Denham said following release of civil parish data and the Super Output Area data for Key Statistics and Univariate tables, no more information would go on NeSS as Beyond 20/20 can only handle so much. He also confirmed that Census Access would wind up in March. There was some cash left in the budget but it was possible that ONS might be required to start charging for supply of data, though the data itself would remain free of charge.
3.4.6
Chris went on to say that ONS are looking at ways of improving the ONS website. This would include providing better links to the 'table finder' which assists users in searching and locating individual files. A new Census home page would also be introduced which would be more user friendly and have better background information with links to the Communication Strategy and Help functions. The prospectus would also be replaced in the next couple of months with a guide containing links to data.
3.4.7
Frank Thomas said that GROS are also currently looking at ways to improve the GROS website also.
3.5
Commissioned Output
3.5.1
Pete Large started this section by saying that ONS had been producing commissioned output since late last year and that they had a backlog of around 60 requests outstanding. The urgent requests were being dealt with first.
3.5.2
Eileen Howes asked when she could expect her data as she had a number of requests outstanding. Pete Large replied that Sophie Boorman was in charge of commissioned output and could provide an update on individual requests.
3.5.3
In response to a question from Richard Price, Frank Thomas replied that it was not possible to obtain daytime population data by postcode.
3.5.4
Brian Green said that NISRA had been dealing with commissioned output requests since July and were currently up to date, although difficulty had been experienced attempting to handle these and the creation of Area Statistics at the same time. He raised the question regarding the policy of respective offices including commissioned output on their web sites. In discussions it was concluded that such a move would be welcomed by users and it was agreed that each of the three offices would separately consider the issue further with a view to implementation.
3.5.5
Frank Thomas said that GROS had received over 70 requests since last March, some of these single requests requiring dozens of tables. At present GROS had about 6 outstanding requests. Frank also said that some commissioned output had been presented by the Scottish Executive to a forum on ethnic classification, proof that commissioned output was very useful to users.
3.5.6
Eileen Howes asked about the charging policy, making the point that at present the person who originally requests the data has to pay for the time taken to put it together, while a subsequent request for the same data would get the data for free. Frank Thomas confirmed that was the case in all 3 Census Offices, but that users could communicate in advance of data requests in an attempt to share the costs. Eileen said that she realised this and that some users were happy to provide the data free to other users themselves. Nick Price said that might be harder to achieve in the private sector.
3.5.7
Frank introduced the subject of SARs (Sample of Anonymised Records) and said that not a lot of progress had been made. Angela Dale from University of Manchester had received SARs as originally specified and these were being checked for population uniques by her team and ONS, but that progress was slow. The SARs would then be made 'safe' by a mix of recoding variables and perturbing individual records. Consideration was being given to enabling certain selected users to get access to 'unsafe' SARs via a Safe Data Lab at ONS in advance of checking being complete, as there was a danger that SARs made 'safe' might not be as valuable to users as first thought.
3.5.8
David Martin said that this hybrid of approaches was probably the best approach but that it was unfortunate that we are still at the stage where data was being produced late as this had an effect on research projects.
4
Geography
4.1
Chris Denham raised the topic of providing old statistics for new areas, saying that lookup tables for 2004 wards would be available from 1st April. Chris invited feedback on whether it would be best to re-run CAS tables for 2004 wards rather than aggregate Output Areas with resulting accumulation of small cell adjustments. Chris also said while a decision had still to be made, ONS would be likely to release Super Output Areas for Key Statistics and Univariate tables.
4.2
Chris then confirmed that Key Statistics and Univariate data for 2004 wards would go on NeSS.
4.3
Richard Price asked whether a 2001 Output Area lookup was available for current postcodes. Pete Large said he would check with ONS Geography. Frank Thomas confirmed that GROS' current postcode index product contains a 2001 Output Area column.
4.4
Frank Thomas said that GROS had 15 geographical areas for which statistics were produced as standard. The database also had Local Enterprise Companies (LEC) areas, Urban/Rural areas based on the Scottish Household Survey definition, Datazones (equivalent to Super Output Areas in ONS & NISRA terms) and NUTS areas. He said that GROS would try to add these geographies to the 1991, 1981 & 1971 databases.
4.5
Brian Green mentioned that NISRA were currently considering the whole issue of Super Output Areas with particular emphasis on what data would be produced (e.g.Key Statistics) and how it would be disseminated.
4.6
Richard Price asked about the possibility of including details of wards actually used in output (2003 wards) as part of an existing ONS Geography product listing inter-censal changes. He agreed to e-mail Chris Denham to spell out exactly what he was looking for so.
4.7
Chris Denham said that ONS hope to issue figures for urban areas, under the ODPM (Office of the Deputy Prime Minister) definition, in a report due out in May.
4.8
Keith Dugmore said that Ordnance Survey future new build plans might provide statistics for new build areas.
4.9
Jenny Boag said that Scottish Local Authorities had been consulted on Output Area to Datazone lookups and that boundaries had been provided. The proposed Datazones however were not always good enough for local purposes and so some authorities with resource and expertise re-drew them, while others did nothing or very little. She confirmed that each Datazone contained between 500 and 1,000 households.
4.10
Chris Denham said that levels 2 and 3 of Super Output Area consultation was underway, led by Robert Heyward at NeSS.
5
Evaluation
5.1
Lucy Baker introduced her paper (OWG(04)01) and asked that all members provide comments by end March.
5.2
Richard Price said that the 2001 Census had gone well although some problems had been experienced as too much of a squeeze had been put on the output phase which affected the quality of the data. He also said that late changes to disclosure control policy were unfortunate. More use of the OWG members could have made a difference, a good example of this being where he and Hywel Davies had helped ONS with the csv file format.
5.3
Eileen Howes added that the GLA had offered to check data during 2001 and was happy to renew this offer. She also said that the late change to disclosure control had been a problem and that it would have been preferable to apply this only to the databases before tabulating, as was the case in Scotland.
5.4
David Martin said that it was essential that decisions on fundamental UK policies such as disclosure control are made and not subject to late change. He recognised that policy changes may be made in non-fundamental areas and that there could well be differences between the Offices in such policy, but that decisions on the main policies should be made early and where possible be consistent across the UK.
5.5
Jenny Boag said that with devolution it could be difficult to keep a common UK approach. Jenny thought that while OWG members should continue to be consulted on 2011, it was just as important to ensure that more people than just the 'experts' (such as OWG) were involved as this could help ensure no late changes were made with policy in 2011.
5.6
Brian Green mentioned that clearly a considerable volume of information had been released from the 2001 Census and enquired of members if it was all being used. Eileen Howes replied that while it was unlikely that any single user was using all of the data, it was highly probable that all the data was required to meet collective user needs.
5.7
Ian White introduced his paper (OWG(04)02) regarding responses received from users to 2011 design and thanked all users for their contributions. Ian confirmed that a compendium of the responses would be published on the ONS website, together with the revised design document, later in March.
5.8
Ian continued by saying that user consultation on the census might not be as self-contained as before since corporate consultation structures were likely to be introduced within ONS into which the traditionally Advisory Groups might have to be subsumed. Ian said that there had to be a commitment to earlier planning for 2011 in order to allow for outsourcing contracts to be negotiated and put in place in good time. This would require close working both with users and within the new divisions within ONS. The present research phase of planning would culminate in a Census Test in England and Wales in 2007.
5.9
Ian confirmed that a round of Advisory Group meetings had been provisionally scheduled for May/June. He noted that the focus of these would be on the 2011 Census and related developments. He expected remaining concerns about 2001 outputs to be covered by OWG, which, he thought, should continue to meet until the program of outputs was complete.
5.10
Chris Denham asked whether consultation needs driven by Census Offices to say that what we are planning or to leave it all open for discussion. Keith Dugmore replied that is should take on the basis of continual improvement.
5.11
Jenny Boag said that the preparation of business cases for Census topics was time consuming but still probably required. Business cases could perhaps be presented with little change for 2011.
5.12
Keith Dugmore said that OWG should look at all statistics, whether they were required for the Census or for NeSS. For example, a question on religion should be discussed with all outputs in mind.
5.13
Ian White noted that the ONS corporate view was that the 2011 Census Team (now part of Designing Sources) does not necessarily have to lead on consultation on topic requirements. However, he felt that it was important to continue direct communication with users to ensure that all potential requirements for outputs feed into the consultation process. It would be useful, therefore to keep OWG meetings ongoing through to 2011.
6
Future of OWG
6.1
Following the previous discussion it was agreed that there would be merit in continuing with OWG meetings in the future, although the exact context was clear at this stage.
7
Any Other Business
7.1
Roger Morgan asked when the remaining CAS tables would be released. Pete Large said that this would be taken care of in the rationalisation exercise about to get underway and that a list of tables that have still to be produced would appear in Census News.
8
Date of next meeting
8.1
A date was not set given that the context of future meetings has still to be decided, although the autumn of 2004 was seen as a possibility.
Action Points
Frank Thomas to ask Alan Fleming to provide test data for Origin Destination to Hywel Davies, Oliver Duke-Williams and Richard Price.
Pete Large to supply geography diagram to Nick Walker and Richard Price to help demonstrate situation on postcode sectors.
Pete Large to inform David Greeno on latest position regarding Standard tables for Health Board areas in England & Wales.
Pete Large to make arrangements with Eileen Howes on supply of MSTR tables.
Pete Large to check position of finalisation and release of 'Focus on' reports and inform David Greeno.
Pete Large to check with ONS Geography whether 2001 Output Area lookup is available for current postcodes and reply to Richard Price.
Richard Price to e-mail Chris Denham outlining request for information on wards at time of the census and wards used for the census as part of lookup table supplied by ONS Geography.
Chris Denham to arrange for next Census Update to include list of tables still to be released.
All members to provide comments to Lucy Baker by end March on paper OWG(04)01.
Census News
