1. Methodology background

 National Statistic   
 Survey name 
 Frequency  Quarterly and annual
 How compiled  Sample based surveys and administrative sources.
Further information on surveys and administrative
sources can be found in How the output is created.
 Geographic coverage  UK
 Sample size 
 Last revised  22 February 2017

Back to table of contents

2. Important points about GDP data

  • GDP measures total domestic economic activity.
  • GDP estimates use 3 approaches; output, expenditure and income.
  • Data for GDP estimates are sourced from survey and administrative sources, which are used in the compilation of individual components of GDP.
  • Data are available and comparable on an annual basis dating back to 1948 and on a quarterly basis dating back to 1955.
Back to table of contents

3. Overview of the output

Gross Domestic Product measures total national economic activity and can be measured in 3 different ways: the output approach, the expenditure approach and the income approach.

GDP estimates are produced quarterly and annually. There are 3 publication stages for the quarterly estimates: the Preliminary Estimate, the Second Estimate of GDP, and the UK Quarterly National Accounts.

The UK national accounts provide the basis for analysing the economic performance of the country and are used throughout business and research communities, education, media and the general public. The accounts are major inputs to HM Treasury (HMT) and Bank of England decisions on fiscal and monetary policy and the forecasts produced by the Office for Budget Responsibility (OBR).

Further information on GDP and the 3 approaches can be found in the National Accounts Methodology and UK National Accounts - a Short Guide.

Back to table of contents

4. Output quality

This document provides a range of information that describes the quality of the data and details any points that should be noted when using the output.

We have developed Guidelines for Measuring Statistical Quality; these are based upon the 5 European Statistical System (ESS) quality dimensions. This document addresses these quality dimensions and other important quality characteristics, which are:

  • relevance
  • timeliness and punctuality
  • coherence and comparability
  • accuracy
  • output quality trade-offs
  • assessment of user needs and perceptions
  • accessibility and clarity

More information is provided about these quality dimensions in the following sections.

Back to table of contents

5. About the output


(The degree to which statistical outputs meet users’ needs.)

The UK National Accounts are compiled in accordance with the European System of Accounts 2010 (ESA 2010), under EU law. ESA 2010 is itself consistent with the standards set out in the United Nations System of National Accounts 2008 (SNA 2008). Significantly, gross national income (GNI), partially derived from the GDP estimates, determines the UK’s contribution to the EU budget. The national accounts cover the UK as a whole, with 3 estimates of GDP published each quarter. In the first month after the end of the reference quarter, the GDP preliminary estimate, based on output, is published. In the second month, these estimates are updated within the Second estimate of GDP release as more detail is available. In the third month, the UK Quarterly National Accounts are published, which includes a full national accounts dataset. These estimates are again updated in the Blue Book, where a fully balanced dataset is published. Further detail on the balancing process is provided in the section ‘How the output is created’.

Regional components of the national estimates are available. These are model-based, or derived from surveys that do not give sufficient sample sizes at smaller areas for reliable estimates to be derived. We publish an annual statistical bulletin presenting regional gross value added (GVA) estimates for English regions, Scotland, Wales and Northern Ireland which include component totals and industry group totals.

Data for GDP estimates are sourced from survey and administrative sources, which are used in the compilation of individual components of GDP.

For information on how we have engaged with users of GDP data, please refer to the ‘Assessment of user needs and perceptions’ section, which is located under the ‘Other information’ heading.

Timeliness and punctuality

(Timeliness refers to the lapse of time between publication and the period to which the data refer. Punctuality refers to the gap between planned and actual publication dates.)

GDP estimates are produced on both quarterly and annual bases. Quarterly preliminary estimates of GDP are published around 25 days after the end of the reference period. The Blue Book consistent dataset is published either in June or September, 6 or 9 months after the reference period.

The Special Data Dissemination Standards page on the International Monetary Fund’s (IMF) website provides more information on periodicity and timeliness of estimates. To date, the GDP statistical bulletins have always met the pre-announced publication dates.

For more details on related releases, the UK National Statistics Publication Hub is available online and provides 12 months’ advance notice of release dates. In the unlikely event of a change to the pre-announced release schedule, public attention will be drawn to the change and the reasons for the change will be explained fully at the same time, as set out in the Code of Practice for Official Statistics.

Back to table of contents

6. How the output is created

We produce a comprehensive set of national accounts of which the main components are:

  • quarterly and annual GDP estimates
  • quarterly and annual sector and financial accounts
  • financial and non-financial balance sheets

The production and publication of each GDP release is managed by a highly skilled team with a strong emphasis on statistical, analytical and economic debate throughout the production process to publish the headline GDP estimate and components.

Trade statistics, Balance of Payments, public sector accounts and other short-term indicators of economic activity are all integrated within the system of national accounts.

It is important to emphasise that the national accounts are estimates of an underlying reality, based on statistical surveys, forecasts and models; they are not compiled through “accounting” in the common sense of the word.

GDP estimates use 3 approaches, which are theoretically equal:

  • output (or production)
  • expenditure
  • income

The national accounts are drawn together using data from many different sources. This ensures that the national accounts are comprehensive and provide different perspectives on the economy, for example, sales by retailers and purchases by households.

The output approach is based upon the following data sources which feed into the Index of Production (IoP), Index of Services (IoS), Output in the construction industry, Retail Sales Index (RSI) and, as a result, the output approach to GDP:

  • data from our surveys including the Monthly Business Survey (this information is gathered directly from businesses) and Construction Output Survey
  • short run forecasts and models where data have not yet been collected or are not yet available
  • our other outputs
  • administrative data

The estimates are then deflated by a variety of sources including lower-level components of the Consumer Price Index, Producer Price Index and Services Producer Price Index.

The expenditure approach is based upon the following data sources which feed in to consumer trends, business investment, UK trade, public sector finances and, as a result, the expenditure approach to GDP:

The estimates are then deflated by a variety of sources including lower-level components of the Consumer Price Index, Producer Price Index and Services Producer Price Index.

The income approach is based upon the following:

These data are subject to many layers of vigorous quality assurance, from clarity and confirmation of individual unit data direct from the business contact to scrutiny of data at the macro level. Other sources of data include other government departments and administrative data, including Value Added Tax (VAT) data from HM Revenue and Customs (HMRC) which are subject to quality checks and challenges from us. By comparing and contrasting these different sources, the national accounts produce a single picture of the economy which is consistent, coherent and fully integrated.

In order to implement improvements reflected in the European System of Accounts 2010 (ESA 2010), we have introduced a new survey to collect purchases data, and have published an article along with a high-level project plan. Purchases data should be updated periodically to reflect the changing composition of businesses’ intermediate consumption in various industries. The Eurostat Manual of Supply, Use and Input-Output Table1 recommends that benchmarked supply and use tables are produced at least every 5 years based on updated source data.

In the past, we have conducted a purchases Inquiry to capture the data required on intermediate consumption. However, at present, the most recent period for which complete and quality assured data are available is 2004, as the survey was suspended, in the main, to reduce costs and compliance on businesses. However, due to strong user demand and to meet legal requirements, we have concluded that the assumptions based on the 2004 data should be brought up to date, by setting up a new purchases survey.

The current situation means that estimates of industries’ intermediate consumption by product are only partially updated and that the UK National Accounts do not meet the EU best practice criterion. The initial unbalanced intermediate consumption matrix has been an area identified that requires development, particularly since the 2008 to 2009 economic downturn.

The initial requirements of a purchases survey, to provide estimates to national accounts, are:

  • an annual survey of current price business expenditure in £000's, for at least 2 (probably 3) years to allow patterns to be verified
  • sufficiently detailed data collection to populate a 114 by 114 industry-by-product matrix
  • delivery of provisional data in December 2016 (reference year 2015), with final data to follow in February 2017
  • a move for most, if not all, industry groups to a biennial or possibly even a triennial survey, from year 4 onwards
  • aggregate purchases to be constrained to Annual Business Survey (ABS) totals
  • no more than 25% of expenditure uncategorised
  • the collection of data for Great Britain initially, and extending data collection to Northern Ireland at a later date

The priority given to introducing a new survey to collect purchases data in part reflects the likely effects on the quality of GDP statistics as purchases data will be used as part of the annual supply use balancing process for GDP. However, it is important to note that GDP is constructed from a wide variety of data sources which seek to maximise the quality of the estimates and as such it is difficult to estimate the current impact of not having more recent purchases data.

Reintroduction of the purchases survey will improve the current estimates of industries’ intermediate consumption by product in accordance with ESA2010 and Balance of Payments and International Investment Position Manual (BPM6) requirements. This will help ensure that the UK National Accounts meet the EU best practice criteria and will also serve to meet strong user demand for up-to-date purchasing patterns for industries. The Purchases Survey aims to collect information on the intermediate consumption of businesses, that is, the goods and services inputs used up or transformed during a production process.

There are different methods for annual and quarterly estimation with subsequent integration. There are 2 main principles driving GDP estimation in the UK:

  • principle 1 – The level of GDP is best estimated using a supply use framework; this breaks the economy down to display transactions of all goods and services between industries and final consumers in the UK, enabling GDP to be based on a fully consistent set of components
  • principle 2 – Short-term growth is best estimated using the production approach to GDP and the other 2 approaches are brought in line with the production approach through careful analysis of the coherence of the 3 approaches
Back to table of contents

7. Production stages

The main stages of the GDP production process are outlined below.

M1 Preliminary estimate

This is based on information on output (production) only and is published around 25 days after the end of the quarter.

This preliminary estimate is based on 44% “actual” data. The rest is based on projections using a variety of modelling techniques. Data are defined as “actual” when they are based on sufficient survey response to produce robust estimates. The percentage of “actual” data is calculated by weighting the data content of each component according to its gross value added (GVA) weight. The data content of the preliminary estimate varies by industry.

M2 Second estimate

The second estimate is published around 7 and a half weeks after the end of the quarter.

In this release, we improve on the preliminary estimate by including more complete output data, as well as providing early information on GDP measured by the expenditure and income approaches. At this point the output approach to GDP is based upon 80% of “actual” data and for the income and expenditure approaches around 50 to 60% of data content is available. The output approach is thought to be the best indicator of growth in the short term. However, at this point, any conflicting information from the expenditure or income sides would be used to inform the average estimate of GDP.

M3 UK Quarterly National Accounts

The third estimate is published 90 days after the end of the quarter.

In this release we produce a full set of quarterly economic accounts, updating and expanding the information made available in the earlier estimate as well as updating estimates for earlier quarters in the current year and normally the previous year. Fuller survey data for components of each of the expenditure, output and income approaches are available. At this point the output approach to GDP is based upon 91% of “actual” data. There is also around 90% data content available to produce estimates of GDP from the expenditure approach and around 70% data content from the income approach. The output approach is still taken to be the best estimate of short-term growth, although again, the other approaches are used to inform the average measure of GDP, as well as to construct the sector and financial accounts.

Annual Blue Book

Annual estimates are published in the Blue Book, usually in July or October. Blue Book is an annual estimate; however, we must emphasise that it is not the first annual estimate, and more so a process that is used for annual reconciliation.

The quarterly data are updated again during the production of the first and second estimates of annual GDP, as data from new and more comprehensive annual data sources become available. The second time an annual estimate is published in the Blue Book, supply and use balancing is applied to the estimate for the first time. The supply and use balancing is re-run in subsequent Blue Books using further benchmark data.

UK Quarterly National Accounts will contain data consistent with the coinciding Blue Book in either June or September.

Methodological improvements may also be made during the publication of the Blue Book; Impact of changes in the national accounts and economic commentary for 2016 describes the changes introduced at Blue Book 2016.

Balancing process

The 3 measures become coherent in the long term through the use of a supply use framework. This enables differences between the estimates of supply and use of specific products to be investigated, and the accounts adjusted to ensure a balance.

We publish information on the methods for balancing the output, income and expenditure approaches to measuring GDP.

In the short run, there are not enough data available to produce a full supply and use balancing table. The first step in increasing the coherence of the raw data received is adjustment for quality by national accounts experts following comprehensive analysis and investigation of possible incoherencies.

Estimates of quarterly growth from the expenditure and income sides are brought into line with the estimate measured from the output side using an alignment adjustment. The output approach is taken to be the best estimate of growth in the short term. The alignment adjustment is applied to the component of the accounts that is conceptually the most difficult to measure and which has the suspected lowest accuracy on a quarterly basis. It is applied to the series “changes in inventories” (on the expenditure side) and ”gross operating surplus of private non-financial corporations” (on the income side).

The size of these alignment adjustments is one measure of coherence of the accounts, and is published in the Quarterly National Accounts and the Second estimate of GDP. These alignment adjustments sum to zero annually as output is not thought to be the best estimate of annual growth because unlike expenditure and income, output does not feed into the supply and use framework which is used to balance GDP.

Further to the alignment adjustments, a statistical discrepancy remains between the 3 approaches until supply and use balancing is run and this is also published. This is the difference between the sum of the expenditure components and average GDP and is published likewise for income components. These are detailed in Table M of the Quarterly National Accounts and second estimate.

The residual error is the amount by which the expenditure-based approach to measuring GDP exceeds the income-based estimate. It is also the sum of the statistical discrepancy (expenditure) with sign reversed and the statistical discrepancy (income) with natural sign (Table L of the Quarterly National Accounts statistical bulletin).

Although a limited audience have access to GDP data ahead of publication, those involved in the process are selected to ensure each GDP balance achieves a rigorous statistical and economic challenge. A “balancing meeting” is held during each production round, where presentations assess GDP and its components against a range of external indicators and a focus on GDP headline components. The data is challenged to ensure consistency and plausibility of the GDP balance. We recognise the importance of transparency and have recently introduced an additional section in our GDP statistical bulletins where the balancing adjustments applied – size and the components targeted – are now published.


Nominal GDP (GDP in current prices) gives the value of GDP at a specific point in time. Growth in nominal GDP reflects the effects of inflation, as well as real GDP growth; it reflects changes in value terms. Real GDP (GDP chained volume measures) excludes any inflationary issues and reflects the changes in volume terms. Using chained volume measures makes use of more up-to-date weights and is therefore more relevant. We have published an article which shows the chain-linking methods in UK National Accounts.

Seasonal adjustment

The headline estimates of quarterly GDP are seasonally adjusted (non-seasonally adjusted versions are available in the United Kingdom Economic Accounts). Seasonal adjustment is the process of removing the variations associated with the time of year, or the arrangement of the calendar, from a time series.

GDP estimates, as for many time series, are difficult to analyse using raw data because seasonal effects dominate short-term movements. Identifying and removing the seasonal component leaves the trend and irregular components.

Further information

We have produced A Short Guide to the UK National Accounts, which gives further information on the content of the national accounts covering uses, principles and compilation.

Back to table of contents

8. Validation and quality assurance


(The degree of closeness between an estimate and the true value.)

Some common pitfalls in interpreting series are:

  • expectations of accuracy and reliability in early estimates are often too high
  • early estimates are based on incomplete data
  • revisions are an inevitable consequence of the trade-off between timeliness and accuracy

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.

Unlike many short-term indicators that we publish, there is no simple way of measuring the accuracy of GDP. 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 estimate of GDP, however, is currently constructed from a wide variety of data sources, some of which are not based on random samples or do not have published sampling and non-sampling errors available. As such it is 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 GDP.

One dimension of measuring accuracy is reliability, which is measured using evidence from analyses of revisions to assess the closeness of early estimates to subsequently estimated values. Many users try to minimise the impact of uncertainty by using the historical experience of revisions as a basis for estimating how confident they are in early releases and predicting how far and in what direction the early release might be revised. The estimate is subject to revisions as more data become available, but between the preliminary and third estimates of GDP, revisions are typically small (around 0.1 to 0.2 percentage points), with the frequency of upward and downward revisions broadly equal.

Many different approaches can be used to summarise revisions; this paper analyses the mean average revision and the mean absolute revision for GDP estimates over data publication iterations. In addition to this analysis, Section 11 of the revisions to gross domestic product in Blue Books 2016 article updates the metrics used to test revisions performance in order to answer the question “Is GDP biased?”.

Revisions are an inevitable consequence of the trade-off between timeliness and accuracy. It is our role to produce the best possible estimate of GDP using all of the available information at that time. Therefore the only way to avoid subsequent revisions to GDP as more information becomes available would be to either delay publication until all the relevant information has been received, which could be up to 3 years after the reference period, or to publish a first estimate and then ignore any subsequent new data and any methodological improvements. So revisions should be treated as generally a good thing, as long as we document the reasons for them and communicate this to users. The balance between necessary revisions and revisions for minor issues is achieved through a published National Accounts Revisions Policy. The results of revisions analysis are regularly presented in the GDP statistical bulletins within the revision triangles and real time databases.

Many different approaches can be used to summarise revisions. The first way to analyse revisions is to look at the simple mean (arithmetic) average revision for estimate of GDP for the period T, between the maturity T+I and the maturity T+j. Figure 1 presents the mean revision by stage of the GDP compilation process, using 10 years’ worth of data.

To remove the effect of offsetting positive and negative revisions giving an average close to zero it is informative to look at the average of the absolute level of revision. The second revision chart, Figure 2, presents the mean absolute revision by stage of the GDP compilation process, using 10 years’ worth of data.

It can be observed that there are more upward revisions than downward; this is partly due to the move in 2011 from Retail Price Index (RPI) to Consumer Price Index (CPI), as the main source of deflation index for the expenditure approach. The total mean absolute revision is relatively sizeable; another factor that will impact this would be that of the changes in measurement during the downturn). More information on the revisions to GDP can be found in Revisions to gross domestic product in Blue Book 2016.

It is important to note that there are other aspects to accuracy, which revisions analysis cannot attempt to measure. A value can be reliable (as in not revised) without being accurate. Broader ways of examining accuracy are presented in Accuracy assessment of National Accounts statistics (2002). The article describes how basic “raw” data are transformed by a series of adjustments to give the statistical estimates that are used to compile the national accounts.

Accuracy of the short-term estimates of GDP growth can be affected by response rates to important surveys. If a lower response rate than normal is received then there is a decrease in the information base of the estimate in the short term, and this may possibly lead to an increased chance of revisions in subsequent estimates of GDP.

Monitoring the Quality of the National Accounts (2008) looks at ways of assessing accuracy, amid a wider discussion of quality in the current climate.

We are continually working on the methodological changes to improve the accuracy of the national accounts.

The first improvement was the harmonisation of the deflators used across the accounts. The second was the replacement of Retail Price Index (RPI) series with Consumer Price Index (CPI) series in forming the deflators. For more information see Deflation improvements in the UK National Accounts (2011). Blue Book 2014 saw an introduction of the European System of Accounts 2010 (ESA 2010) and a series of articles were published explaining the Changes to National Accounts.

Accompanying each quarterly and annual production cycle, external quality assurers with particular areas of expertise are invited to challenge and report on the statistical and economic coherence of the headline national account and component dataset. Current assessors include HM Treasury, Bank of England, National Institute of Economic and Social Research, HM Revenue and Customs and Tax Administration Research Centre. The external quality assurers work to challenge the synergy of the dataset from a full range of views – those of producers, data compilers and users of the statistics – before final sign-off.

Coherence and comparability

(Coherence is the degree to which data that are derived from different sources or methods, but refer to the same topic, are similar. Comparability is the degree to which data can be compared over time and domain for example, geographic level.)

Since international standards such as SNA 2008 and ESA 2010 are used in the production of the national accounts, the figures should be directly comparable with the accounts of other countries. However, the revisions policies of these countries should be examined before comparing data for back periods.

Data in the GDP statistical bulletins are consistent with our following outputs:

The only inconsistencies occur when more timely monthly releases introduce revisions in advance of their incorporation into the later quarterly publications. For example, data are not always consistent with trade in goods as the UK trade release contains more up-to-date quarterly data.

Inconsistencies with the Public Sector Accounts releases are also possible due to the different revisions policies being applied to these releases.

When annual data first become available (in the February M2 release) they contain revisions to previous quarter data. This introduces inconsistencies between the latest Second estimate of GDP data and sector and financial accounts data published in the earlier Quarterly National Accounts M3 release.

Another important aspect is the coherence between the 3 different approaches to the measurement of GDP, which are theoretically equal. However, since they are measured independently, statistical and measurement errors will mean that this is not the case. Further explanation is provided within the ‘Balancing process’ section of this document. These issues also formed a major part of the article Monitoring the quality of the National Accounts (2008).

Every effort is made to ensure that the series is comparable over time, and a comparable time series is available back to 1948 for annual estimates (1955 for quarterly estimates). Where possible, changes to methodology are applied to the whole series to ensure this comparability is maintained. However, the National Accounts revision statements may mean that this is not possible. For more information see Methodological Improvements, ESA10 changes to historic GDP for BB14.

Back to table of contents

9. Concepts and definitions

(Concepts and definitions describe the legislation governing the output and a description of the classifications used in the output.)

GDP estimates are produced in line with international standards, most notably ESA 2010 which is enforced for all EU Member States through EU council regulation (EC) No 549/2013. ESA 2010 is in turn consistent with the United Nations System of National Accounts 2008 (SNA2008). The SNA 2008 is the recent update of SNA1993 and led in turn to the revision of ESA 1995 forming ESA 2010, which has subsequently been implemented in the UK National Accounts, as well as those of all other EU member states. We also provide further information on the introduction of ESA 2010.

GDP estimates are compiled using UK Standard Industrial Classification 2007 (SIC 2007), used for the first time at Blue Book 2011, replacing the previously used SIC 2003. The introduction of SIC 2007 for GDP estimates was in keeping with EU regulations and adapted the classifications to changes in the structure of the economy. Important changes in SIC 2007 include a number of new sections giving more service sector detail while the detail in manufacturing is significantly reduced, reflecting the move towards more services-based economies over the past 20 years. For further information on the introduction of SIC 2007 see Implementation of Standard Industrial Classification 2007: December 2009 update, Economic and Labour Market Review, December 2009, Volume 3, No. 12.

Back to table of contents

10. Other information

Output quality trade-offs

(Trade-offs are the extent to which different dimensions of quality are balanced against each other.)

There is a trade-off between accuracy and timeliness. Provisional outputs are timely, but less firmly based.

Estimates may be revised during intermediate stages. These are explained under the section ‘How the output is created’.

Assessment of user needs and perceptions

(The processes for finding out about users and uses, and their views on the statistical products.)

User engagement surveys for the second estimate of GDP and the Quarterly National Accounts were conducted from May to July 2015. In accordance with the Code of Practice for Official Statistics requirements, the objectives of the user engagement surveys were to investigate:

  • who the users of the statistical product were
  • what the statistics were used for (including the decisions they informed)
  • users’ perceptions of the quality of the statistics, statistical presentation and statistical commentary
  • users’ perceptions of the statistical service in relation to this particular statistical product

The results of the Quarterly National Accounts user engagement survey and Second estimate of GDP user engagement survey are available.

We plan to run the user engagement surveys every 1 to 2 years.

Back to table of contents

11. Sources for further information or advice

Accessibility and clarity

(Accessibility is the ease with which users are able to access the data, also reflecting the format in which the data are available and the availability of supporting information. Clarity refers to the quality and sufficiency of the release details, illustrations and accompanying advice.)

Our recommended format for accessible content is a combination of HTML web pages for narrative, charts and graphs, with data being provided in usable formats such as CSV and Excel. Our website also offers users the option to download the narrative in PDF format. In some instances other software may be used, or may be available on request. Available formats for content published on our website but not produced by us, or referenced on our website but stored elsewhere, may vary. For further information please refer to the contact details at the beginning of this document.

For information regarding conditions of access to data, please refer to the links below:

In addition to this Quality and Methodology Information document, basic quality information relevant to each release is available in each GDP statistical bulletin. Advance notice of any forthcoming major changes in methodology for the GDP estimates can be found under National Accounts methodology and articles.

Useful links

Guide to Gross Domestic Product: Measuring the UK's economic activity UK National Accounts - a Short Guide.

Back to table of contents