This section provides guidance on planning, pre-announcing and withdrawing statistics publications. We have included a PDF version of this guidance at the bottom of the page for use offline and printing.
It is important that users, stakeholders and providers are consulted on plans for the statistics that we produce. Additional guidance relating to user engagement can be found in the User Engagement Guidance pages.
Determining Publication Dates
The primary considerations in determining the month of publication date are mostly encompassed by principles and practices of the Code of Practice for Official Statistics and the Quality Assurance Framework for the European Statistical System (ESS QAF)
- user needs (pillar V1, principle V1.1) and relevance (ESS QAF) - when are the statistics needed for the key uses to which they will be put. It is entirely reasonable to consider specific requirements such as informing major policy decisions, providing evidence in advance of or in association with publication of key policy documents, providing an informed basis for Parliamentary debates. It is not helpful to anyone for statistics to be released the day after a date when they could have had a significant impact.
- timeliness (ESS QAF) - minimising the gap between the occurrence that is being measured and the published output
- releasing statistics as soon as they are judged ready (pillar T3, principle T3.5) with adherence to pre-announced dates (pillar T3, principle T3.1)
- orderly publishing (pillar T3) and coherence (ESS QAF, pillar Q3) - ensuring that the statistics give a sufficiently full and coherent picture - implying that statistics are grouped together where need be and not released out of context
- standard dates for quarterly or more frequent outputs
- other ESS QAF considerations including accuracy and reliability, coherence with other outputs (e.g. if it makes sense to release statistics across the UK at the same time or alongside other reports e.g. research reports)
- risk management and contingency planning - having to delay publications due to unforeseen problems arising can undermine trust. Risks should be assessed and appropriate contingencies should be built in
- work planning and prioritising - clearly the above considerations need to be balanced against each other and such judgements need to be made across the whole range of work of the unit relative to available resources.
- publication during recess or at certain other times may lead to issues of trust especially if it is likely that Parliament is likely to want to debate the statistics. However this should only be a factor if it does not impinge on the above principles. It would not be appropriate to delay statistics until after an election. However, where there is sufficient notice of a national polling day, statistics should be scheduled to avoid this date. If an election or referendum is called at short notice, any statistical releases which had been pre-announced to be published on the polling day should be rescheduled to the next day. The Government Statistical Service (GSS) policy on publishing on polling days is available at: GSS Policy on Publishing Official Statistics on Polling Days. Separate guidance will be issued regarding the publication of statistics in a pre-election period. The guidance for the 2016 Scottish Parliament Election is available at: http://www.gov.scot/Publications/2016/01/3207/323681
See Scottish Parliament recess dates.
When determining the exact day of publication the above factors should all be considered further. In addition, it is legitimate to set the date to maximise the impact of the statistics.
Under no circumstances should publication dates be set or changed in expectation that the results will be good or bad news. Final decisions should rest with the responsible statistician, unless otherwise determined by the Head of Profession. Advice may be sought from policy colleagues and communications colleagues. Any political pressure or other inappropriate attempts to influence dates must be reported to the Head of Profession as soon as possible.
Pre-announcing release dates and deciding such dates on an objective basis are key principles of the Code of Practice for Official Statistics (Pillar T3: Orderly release).
There are three levels of pre-release access.
- Statistics that have a tangible effect on the value or volume of trading in markets are the most sensitive. For these statistics, we will not have pre-release access. The paper 'Market Sensitivity of Economic Official Statistics on the Scottish Economy' outlines a method for assessing market sensitivity of statistics and concludes that we don’t currently have any market sensitive statistics in Scotland. However, when we either produce new economic statistics, or bring forward the timing of our existing series this gives a method to test their market sensitivity.
- Short-term economic and public finance statistics are not currently market sensitive, but are still very sensitive. As such, these statistics will have a maximum of 24 hours pre-release access.
- For all other official statistics in Scotland, we will maintain 5 days pre-release access.
Pre-announcement is handled through a central system operated by the Office of the Chief Statistician and Data Officer (OCSDO). This is fed by the forthcoming publications spreadsheet. OCSDO use the information provided in the spreadsheet to produce a report to send to Ministers, to publish on the Forthcoming Publications page of the statistics website.
Publications should be pre-announced in 2 stages:
Full details, including the month of publication should be pre-announced at least 12 months before publication.
To do this, Branches should review their year ahead publications every quarter ensuring that the forthcoming publications spreadsheet includes all publications in the next fifteen months so that there is always a full year of planned publications on the schedule.
The exact date should be published the preceding month (earlier if known with certainty). For market sensitive statistics or those produced on regular, quarterly or more frequent basis, the exact date should be given at least six months ahead where this is known. Again, this should be included on the forthcoming publications spreadsheet.
The pre-announcement schedule is made available on the Forthcoming Publications page on the Scottish Government website.
The information required for this schedule is as follows.
(A) Publication series
(C) Publication date/month
- If you cannot specify the month then leave blank and insert a comment into the cell outlining the details.
- This should be a summary of the main statistics that the publication will cover. This should normally be no more than 20 words and should not repeat what is obvious from the title of the publication.
(F) Contact details
- Frequency (e.g. "Annual", "Quarterly" etc.) where the same statistics are produced on a regular basis and 'occasional' or some other appropriate description for all other releases.
(H) Reason for change
- Fill this out if the date is different from a date previously indicated
- This would also cover any releases that for any reason were not included in previous schedules.
- The reason should always be made clear
- The Head of Profession has final clearance on date changes. Please contact the Head of Profession for advice on changes of publication date.
(I) Publication Type
- For 'Type of Statistics', please indicate whether the statistics are National Statistics, Official Statistics or Experimental Statistics.
Changing Publication Dates
Once the exact date has been announced it should be changed only in extreme cases, and this must be cleared with the Head of Profession as soon as possible.
Please note that changes to dates that have not been given exactly may vary.
The need for any particular set of statistics should be subject to regular review - at least once a year. Withdrawal of statistics is dealt with in much the same way as additions in terms of reporting and consultation.
We will consult with users before withdrawing any official statistics. The consultation should clearly advise users of the intention to withdraw the publication, stating why it is being withdrawn. Ideally it will outline any alternative sources of data available to users, if these are available. The Head of Profession should be advised of these instances as they will liaise with the National Statistician after the consultation, as the product, if it is a National Statistic will require to be de-designated, hence the UK Statistics Authority will also be involved.
Statisticians should avoid withdrawing a publication without user consultation; if in any circumstances this should happen, the responsible statistician should discuss this with the Head of Profession (Chief Statistician).
Timing on day of Release
All statistical releases must go out at 09.30, and no other statements (news release, ministerial statement) should go out any earlier than this. This is a requirement of the Code of Practice under Pillar T3, Principle T3.6.
Other times may be agreed by exception. For statistics that have to be formally laid before Parliament, 09.30 should still be workable, but copies should be supplied to the Chamber desk the preceding afternoon, with clear instructions about the timing of release. Anyone who receives the statistics in advance of their publication should be recorded on the Pre-Release Access list, in accordance with Pillar T3, Principle T3.4 of the Code of Practice.
Experimental Statistics; data being developed
Experimental Statistics are an type of official statistic that are released early in their production to involve users and to build in quality at an early stage. The nature of these statistics means that it may not always be possible or appropriate to publish a detailed release calendar; the Scottish Government has produced specific guidance to cover 'Experimental Statistics; data being developed', and the governance arrangements for their production and publication.