Background notes

We welcome your feedback on this publication. If you would like to get in touch, please contact us via email: blue.book.coordination@ons.gov.uk.

As part of our user consultation, we are conducting a Blue Book survey. The aim of the survey is to find out:

• how you use our statistics

• your understanding of the data published

• your views on the quality of the publication

Your responses will help us improve some of our most important products. We will analyse the responses and publish a summary of the results over the forthcoming months.

The survey will take about 10 minutes to complete and will close on 09 September 2016. All answers will remain anonymous. Our confidentiality statement has further details.

On 16 May 2016 all stakeholders were invited to respond to our draft of the first edition of our Economic Statistics and Analysis Strategy (ESAS) to prioritise and guide ONS’s work on economic statistics, inviting all stakeholders to comment by 30 June 2016. ONS has already produced a strategy for the National Accounts and the ESAS encompasses this and goes wider to cover all economic statistics.

This strategy will be reviewed and updated annually in the light of changing needs and priorities, and availability of resources. Making explicit ONS’s perceived priorities will allow greater scrutiny and assurance that these are the right ones. In addition, this ESAS will allow research and development priorities to be laid out, making it easier for external experts to see the areas where ONS would be particularly keen to collaborate.

The UK national accounts are produced under internationally agreed guidance and rules set out principally in the European System of Accounts (ESA 2010) and the accompanying Manual on Government Deficit and Debt- Implementation of ESA 2010 – 2016 edition (MGDD).

In the UK, we are responsible for the application and interpretation of these rules. Therefore we make classification decisions based upon the agreed guidance and rules.

Common pitfalls in interpreting series:

• expectations of accuracy and reliability in early estimates are often too high

• revisions are an inevitable consequence of the trade-off between timeliness and accuracy

• early estimates are based on incomplete data

Very few statistical revisions arise as a result of “errors” in the popular sense of the word. All estimates, by definition, are subject to statistical “error”. In this context the word refers to the uncertainty inherent in any process or calculation that uses sampling, estimation or modelling. Most revisions reflect either the adoption of new statistical techniques or the incorporation of new information which allows the statistical error of previous estimates to be reduced. Only rarely are there avoidable “errors”, such as human or system failures, and such mistakes are made quite clear when they do occur.

The quality of Blue Book estimates

Unlike many of the short-term indicators we publish, there is no simple way of measuring the accuracy of the Blue Book dataset. All estimates, by definition, are subject to statistical uncertainty and for many well-established statistics we measure and publish the sampling error and non-sampling error associated with the estimate, using this as an indicator of accuracy. Since sampling is typically done to determine the characteristics of a whole population, the difference between the sample and population values is considered a sampling error. Non-sampling errors are a result of deviations from the true value that are not a function of the sample chosen, including various systematic errors and any other errors that are not due to sampling. The Blue Book dataset, however, is currently constructed from various data sources, some of which are not based on random samples or do not have published sampling and non-sampling errors available, making it very difficult to measure both error aspects and their impact on GDP. While development work continues in this area, like all other G7 national statistical institutes, we don't publish a measure of the sampling error or non-sampling error associated with this dataset.

Estimates for the most recent quarters are provisional and are subject to revision in the light of updated source information. We currently provide an analysis of past revisions in the GDP and other statistical bulletins that present time series.

Our revisions to economic statistics page brings together our work on revisions analysis, linking to articles and revisions policies.

Revisions to data provide 1 indication of the reliability of main indicators. Tables 3 and 4 provide a summary on the size and direction of the revisions that have been made to data covering a 5-year period. A statistical test has been applied to the average revision to find out if it is statistically significantly different from zero. An asterisk (*) shows if the result of the test is significant.

An article entitled HMRC VAT project update was published on 12 July 2016, the fifth in a series of articles. This article outlines the latest position of the VAT project, including next steps. Four previous articles have been published in this series:

• HMRC VAT project update, 4 April 2016.

• Feasibility study into the use of HMRC turnover data within Short-term Output Indicators and national accounts, 14 August 2015.

• Exploitation of HMRC VAT data, 7 October 2015.

• "HMRC VAT project update" 21 December 2015.

You can get the latest copies of this and all our other releases through the release calendar on our website.

Details of the policy governing the release of new data are available from the media relations office.

We are committed to ensuring that all information provided is kept strictly confidential and will only be used for statistical purposes. Further details regarding confidentiality can be found on our website in the respondent charters for businesses and households.