The Official Statistics Section held a meeting on 25 June 2013, chaired by RSS president John Pullinger, with Sir Andrew Dilnot, chair of the UK Statistics Authority, as speaker.
Andrew spoke about the Authority’s strategy statement published on 14 February 2013. He discussed the five strategic priorities for the UK Statistics Authority in detail and some of his responses to related points raised under the question and answer session are included below:
- Quality: Including a more selective assessment programme. The selection process would be based on a number of factors such as importance. Poor quality can result from failures at any point along the statistical production chain, from data source to final output. This is why the code of practice and the assessments have been structured so as to examine all of these. Andrew agreed that there were challenges to be faced with increasing use of administrative source data, but GSS statisticians should welcome this.
- Impact: Including more focus on providing explanation of the data to users and navigational improvements to the ONS website. Although the ONS website has improved there is clearly much more to be done. Andrew acknowledged that the National Statistics Publication Hub website could now benefit from a review. Measuring impact is a question being looked at by the Authority.
- Efficiency: ONS is coping well with the decline in the non-census budget, but difficult choices remain. It may be better to stop some activities altogether than ‘salami slice’ everything.
- Coverage: The UK has an excellent range and depth of official statistics. It is the job of the Authority to lead and stimulate debate on priorities.Preliminary results from the project on alternatives to the census indicate that good quality annual population counts could be obtained from a combination of the Health register and the DWP new Citizens Information System (CIS). However it will be more difficult without a traditional census to deliver other local area census outputs such on economic activity of households and housing. Users need to articulate these needs clearly.In response to a question Andrew said that there were insufficient resources at the Statistics Authority to extend its coverage to local government statistical work. However, the principles still apply, including where local authorities are the data source for official statistics.
- Trustworthiness: Opinion research carried out by Ipsos-MORI illustrated the problem. Politicians are not trusted and most high profile public use of official statistics are made by politicians, hence the contagion effect results in low trust for official statistics.
Work is also ongoing on regarding: (i) increasing the number of national statistics subject to no advance release access (now standing at 20 per cent); and (ii) reducing the number of people on the lists of the remaining 80 per cent (which now average 18.2 per press release). Andrew agreed that, although important, this was not his number one priority (which was greater acceptance by politicians of the need for evidence-based decision making) and that advance release dates had had an even more important influence on the credibility of official statistics.
John Pullinger gave his response as a discussant. He agreed with all of these priorities and welcomed the production of a written statement of the Authority’s strategy. He argued that in the modern data economy, official statistics are an essential element of UK national infrastructure. The Statistics Authority Board is the steward of this infrastructure and must build its value on behalf of its stakeholders, UK citizens and decision makers. The Board should focus on identifying the questions on which we as a nation need statistical evidence to make sound decisions and then be accountable to us all for getting it. He pledged support from the RSS community to help bring this about.
Questions from the floor (not covered above)
- Decision making on new data and cuts: The Authority will decide on ONS statistics and have an important influence of decisions by government ministries. It was noted that there is no overall statistics work plan as is often the case in other countries; but that the Authority keeps track of and responds to all major decisions on statistical outputs by ministries. More generally, Andrew had noted an increase in the awareness of the importance of good statistics by senior officials.
- Health Statistics: Due to budgetary constraints, the Health and Social Care Information Centre had withdrawn funding for some analytical work in the area of health statistics, and that it had been necessary for ONS itself to review what it was able to deliver here given the pressures on it financially. Andrew recognised that this impacted on users of health statistics and emphasised that challenging budgetary environment in this and other areas of official statistics. A questioner said that the recent publicity of poor decision making in other health-related bodies such as the Care and Quality Commission illustrated the need for the statistical ‘head of profession’ role to be carried over beyond ministries and into arms-length bodies.
- GSS: Was the separation of statisticians between ONS and ministries getting worse? Andrew did not agree, but he would like to see more movement of professional staff within government and with statistical posts outside of government.
Report by Philip Turnbull, meeting organiser