Commentary on supply and use balanced estimates of annual GDP: 1997 to 2014

This article explains the revisions from 1997 to 2014, between the estimates of gross value added (GVA) and gross domestic product (GDP) published in the 2016 supply and use tables (SUTs) and those previously published in 2015. It also covers the revisions to the GVA and GDP estimates for 2014 derived in the Quarterly National Accounts, which have now been replaced by the first supply and use balanced estimates for that year.

The Quarterly National Accounts use a variety of short-term indicators to form estimates of GDP. Annual supply and use balancing uses more detailed indicators, which only become available after 1 to 2 years. In annual supply and use balancing, the 3 approaches to GDP (income, output and expenditure) are fully reconciled using a matrix of industries and products.

Further information on the latest estimates can be found in a recent article “Impact of changes in the national accounts and economic commentary for 2016 Quarter 1 (Apr to June)”. This article highlights the impacts of the improved methods introduced in Blue Book 2016 and examined the revisions that have occurred to the headline aggregates within the national accounts. In addition, Chapter 1 of Blue Book, 2016 Edition presents a more detailed analysis of the main trends emerging from the newly available annual data across the main published sectors.

Annual current price supply and use balanced estimates of GDP have been published by the Office for National Statistics (ONS), with the exception of 2007, in every year since 1992. For the 2016 edition of the Supply and use Tables (SUTs), annual estimates of current price GDP have been revised from 1997 to 2013 and first estimates have been published for 2014, which replace earlier estimates for that year published in the Quarterly National Accounts (Tables 1a, 1b, 1c and 1d).

Two layers of revisions have been applied to the 2015 Blue Book supply and use estimates to produce the 2016 Blue Book estimates:

• Layer 1 – various methodological improvements, including those introduced to bring about compliance with new international standards and guidelines, changes from ensuring comparability and other changes to meet user needs

• Layer 2 – fully open revisions from 2012 to 2013, including those methodological improvements introduced in Layer 1 but also any revised data

Revisions to 2014 data as shown represent differences between annual data previously estimated in the Quarterly National Accounts and supply and use balanced estimates that have been produced for the first time. These differences therefore include both methodological changes and data differences.

Tables A1.1a and A1.1b present the revisions by main GDP component from 1997 to 2014. Compared with the 2015 vintage, annual growth in GVA at basic prices has been revised down by an average of -0.39 percentage points and in absolute terms (disregarding the sign) by an average of 0.40 percentage points. The revisions to annual growth in GDP at market prices are -0.35 percentage points and 0.36 percentage points respectively.

The first supply and use balanced estimates of annual current price GDP in 2014 indicate GVA growth of 4.7% and GDP growth of 4.8%. These compare with the previously published 2014 growth estimates derived from quarterly balancing of 4.7% for both GVA and GDP growth (Table 1a).

As highlighted in the article “Impact of changes in the national accounts and economic commentary for 2016 Quarter 1 (Apr to June)" the majority of the current price revisions over the period since 1997 comes from improvements to the methods used to estimate imputed rentals, which is the value to home owners of their dwellings. The impact of these (along with other) changes is generally to increase the level of current price GDP, although the growth rate revisions vary for individual years. The upward revisions to levels, from the imputed rental changes, reduce towards the more recent periods but nevertheless this remains a key contributor to the revisions in annual current price GDP growth rates. The broad timing and depth of the downturn in 2008 to 2009 stays the same, and the overall magnitude of the recovery immediately following the downturn remains the weakest of the past half century. The 2008 to 2009 downturn is a little deeper than previously estimated. The fall in output between the peak and trough (Quarter 1 (Jan to Mar) 2008 to Quarter 2 (Apr to June) 2009), is estimated to have been 6.3%. The overall strength of the recovery is largely unrevised, but the timing has shifted slightly. This resulted in UK output regaining its pre-downturn peak in Quarter 3 (July to Sept) 2013, one quarter later than previously estimated.

The remainder of this article is structured as follows:

• Section 3 provides background on the supply and use balancing approach
• Section 4 gives a detailed breakdown of the revisions in each layer
• Section 5 discusses future improvements in the production of supply and use estimates
• Tables A1 and A2 provide supporting information and data tables

Why do we use supply and use balancing for annual GDP estimates?

Supply and use balancing fully reconciles production, income and expenditure data at a very detailed level to produce balanced estimates of gross value added (GVA) and gross domestic product (GDP) in line with European System of Accounts 2010 (ESA 2010). The estimates underpin all but the most recent estimates of GDP, which are based solely on the quarterly approach to balancing. It differs from quarterly balancing in several ways.

Quarterly balancing uses a variety of short-term indicators of economic activity to form estimates of GDP by the production, income and expenditure approaches. Annual supply and use balancing uses a separate, more detailed set of indicators which only becomes available after lags of around 1 to 2 years. In particular, the supply and use process uses the Annual Business Survey which includes estimates of turnover and business purchases that form part of the basis of estimates of output and intermediate consumption. The Monthly Business Survey used in the quarterly process only provides estimates of turnover.

In quarterly balancing, the production information is considered to be the strongest, so income and expenditure levels and growth are mathematically aligned to this via an annual “statistical discrepancy” and a quarterly “alignment adjustment”. In annual supply and use balancing, the 3 approaches to GDP are fully reconciled using a matrix of 114 industries and 114 products, so no constraining adjustments are required.

Quarterly estimates across the 3 GDP approaches are produced rapidly, at 60 days (T+60) and 90 days (T+90) after the reference quarter, but the trade-off between timeliness and accuracy limits their data content and detail. Annual supply and use estimates, though more accurate and detailed, are first published at the earliest 18 months after the last reference year.

Quarterly balanced GVA and GDP are produced in both current prices and chained volume terms and with and without seasonal adjustment. Annual supply and use balanced GVA and GDP are only produced in current prices and as they are annual, no seasonal adjustment is required.

For each Blue Book, the quarterly production, income and expenditure estimates are brought into line with the latest annual supply and use balanced estimates up to the latest supply and use reference year, while the quarterly estimates beyond that are led by the newly-aligned production estimate.

Further detail on the annual and quarterly balancing methods and their interaction can be found in “Balancing the Three Approaches to Measuring Gross Domestic Product, 2012”

How is supply and use balancing carried out?

The supply and use approach, or to give it its full title, the input-output supply and use approach to balancing annual current price GDP ensures that all the components of GDP are fully reconciled, to the nearest £ million, by ensuring that:

• The supply of products – the goods and services produced by the domestic market and non-market sectors plus any imports equals the demand for products by domestic producers and consumers plus any exports

• The inputs to industries – goods and services used up during production plus the primary inputs of labour and entrepreneurship equals the outputs from industries

The approach also takes account of the fact that supply is measured at basic prices while demand is measured at purchasers’ prices. This is done by making adjustments to supply accounting for taxes and subsidies on products and production as well as trade and transport margins.

From the point at which source data becomes available for processing, to the point at which final balanced estimates are signed-off, takes around 6 months. The final estimates must be reached in time for the subsequent benchmarking of the quarterly GDP path and the derivation of the Sector and Financial Accounts for each annual Blue Book, by the end of Quarter 1 (Jan to Mar) for a June or July publication or by Quarter 2 (Apr to June) for a September or October publication.

Large amounts of data are used and produced by the balancing process so a rigorous framework must be adopted to make the process both efficient and robust. There are various different ways of achieving balanced current price outcomes, each with its own merits and issues. A discussion of these can be found in “Input-Output Supply and use Tables – Balanced Estimates of GDP using a Supply and use approach” but the essential differences are whether balancing is carried out sequentially or iteratively and manually or automatically.

The sequential approach aims to establish the strongest elements of GDP first and balance the remaining elements around those levels. For example, GDP by the income approach might be set initially, after which the levels for GDP expenditure and GDP production would be brought into line. Alternatively, the strongest elements of each approach to measuring GDP can be established first for example, the levels of industrial output (production), compensation of employees (income) and household final consumption (expenditure) – and the remaining elements can be balanced in around these afterwards.

In contrast, the iterative approach uses alternating rounds of manual product supply and demand and industry input and output balancing to arrive at final estimates of the three approaches to GDP only at the end of the process. The basis of this approach is that the supply and use of products and industry inputs and outputs are intertwined through the output and intermediate consumption of products, so each product balance unbalances the industry picture and vice-versa. As the number of iterations rises, the product and industry imbalances gradually reduce towards zero. This approach is used by ONS and is described in further detail below.

Before balancing begins, time is spent assessing the raw inputs in their own right and making any necessary quality adjustments. Where necessary, further specific quality adjustments are also made in the preliminary iterations before the general reconciliation process begins.

For each iteration, Supply and Use Production Branch co-ordinates trained staff, each of whom is allotted a specific tranche of the 114 industries or products to balance based on their area knowledge. Balancers are given specific guidance after each product and industry phase from a senior steering group and Supply and Use Production Branch, as well as access to supporting sources of evidence. At the end of each iteration, rebalanced industry and/or product data are collated, quality assured and processed through central systems to be recorded and to produce the next iteration.

Once the annual imbalances are economically insignificant – for example, at less than 0.1% of Total Supply – a “raking and scaling” algorithm is applied to apportion these across the economy and reach the final balanced position. The use of an automatic balancing algorithm as a primary tool has significant drawbacks, in that a given mathematical solution may not account for economic evidence. However, used as a final tool, it is superior to manual balancing as it removes any small discrepancies for which there is no economic justification, both impartially and efficiently.

How are the data published?

Supply and use data are primarily published in extensive, cross-sectional tables to enable the large amount of detail to be seen in the clearest context. Summary tables are also published in Chapter 2 of the Blue Book which combines the 114 industry/product groups into ten sections.

Since 2011, the UK National Accounts have been produced using the Standard Industrial Classification 2007 (SIC 2007) and the Classification of Products by Activity 2008 (CPA 2008), which are consistent with the European standard reference framework for the classification of economic activities (NACE Rev.2). The supply and use industry/product groups are fully consistent with SIC 2007 and CPA 2008 in terms of their coverage, numerical codes and descriptions.

Methodological revisions from 1997 to 2011 (Layer 1)

The UK National Accounts adhere to international standards, set out in the European System of Accounts 2010 (ESA 2010) and the UK Balance of Payments consistent with the International Monetary Fund’s Balance of Payments and International Investment Position Manual sixth edition (BPM6), which ensure comparability across EU member states. These manuals are updated periodically to reflect economic and technological developments in domestic and global economies as well as changes in user needs. In 2014 the UK was required to publish the UK National Accounts consistent with the European System of Accounts and Balance of Payments Manual. These manuals replaced the European System of Accounts 1995 (ESA95) and the Balance of Payments Manual fifth edition (BPM5). The updated standards are in line with the UN System of National Accounts 2008 (SNA2008), which covers all countries on a voluntary basis.

Work focused on implementing improved methods to address gross national income (GNI) reservations currently in place on the National Accounts on an ESA95 basis and the transition to the new standards in ESA 2010 and BPM6.

Gross National Income (GNI) is an important statistic within the National Accounts, and it is used in the calculation of a member state’s contribution to the EU budget. GNI or, as it was previously known, Gross National Product (GNP), describes the total primary income received by residents of a country and links the economic activity described by GDP with the destination of the income so generated. To give a full definition, GNI is equal to GDP plus net property income from abroad. Property income is not (as might be suggested by the name) the income generated by the ownership of buildings (rental). It is in fact made up of interest, the distributed income of corporations (dividends, repatriated profits and so on) and rent on land. This means that countries can have very high levels of GDP, but GNI would be significantly lower if, for example, many of the production units were owned by multi-national corporations with their headquarters in other countries.

Due to the administrative importance of the GNI statistic, the EU statistical office (Eurostat) carries out regular audits of the methods and data used to estimate GNI. In 2012, following a comprehensive audit of the methods used across EU countries, a number of areas for improvement were identified which all member states needed to address. These improvements are known as “reservations”. The UK National Accounts addressed a number of these reservations in Blue Book 2014, and all the remaining outstanding UK reservations were addressed in Blue Book 2015.

This article provides a summary of each of the revisions (ESA10 and other methodological revisions) being implemented in Blue Book 2016.

Tables 2a and 2b show the revisions to GVA and GDP Layer 1 estimates. Tables A1.2a and A1.2b summarise the revisions to the main aggregates applied in Layer 1 from 1997 to 2011.

Methodological changes that have been introduced which have an impact on GDP for 1997 to 2011. Imputed rental is an estimate of the housing services consumed by households who are not actually renting their residence. Conceptually it is the amount that non-renters pay themselves for the housing services that they are producing. This is a legitimate and important component of GDP since the value represented by housing services should not depend upon whether they are consumed by an owner occupier or otherwise. Improvements to the data sources and methods used to estimate imputed rental have been included in the Blue Book 2016 dataset. This includes the use of new data sources (Valuation Office Agency data and similar data from the devolved nations) replacing existing Living Costs and Food Survey data. Further detail can be found in the article Methodological improvements to National Accounts for Blue Book 2016: Imputed Rental.

Exhaustiveness

As part of a GNI ESA 1995 reservation, improvements were implemented to the exhaustiveness adjustment for concealed income in Blue Book 2015, as described in the article Methodological improvements to National Accounts for Blue Book 2015: Exhaustiveness. The revisions for 2008 onwards in Blue Book 2016 are as a result of the annual reassessment of the tax gaps analysis by HM Revenue and Customs (HMRC), which in turn leads to revisions in our estimates of concealed income. There are also some smaller revisions pre 2008 as a result of a reassessment of how best to include concealed income adjustments within the Supply and Use tables, which required a small rebalancing.

Estimates of Value Added Tax fraud

Scrutiny of estimates of non-complicit Value Added Tax (VAT fraud) has led to improvements which are now being implemented. VAT fraud is thought to occur when individuals or corporations who under- or non-report their respective salaries and profits to the tax authorities ("Tax Evasion") also choose to hold onto the VAT that they charge to non-complicit customers on their underreported or unreported activity. It is thought that by paying VAT on these transactions they would reveal the wider evasion. Because current estimates of tax evasion are at basic prices, meaning before the inclusion of taxes on products, this form of VAT fraud is not currently captured in the National Accounts.

Insurance claims by non-profit institutes serving households

As part of addressing a GNI ESA 1995 reservation, improvements were implemented to the measurement of non-profit institutions serving households (NPISH) in Blue Book 2015. Further scrutiny has identified a small inconsistency in the treatment of insurance within the NPISH sector. Previous treatment included the insurance premiums and supplements paid by NPISH within the National Accounts, but not the claims received by the NPISH sector. This has now been completed with minimal impact on current price GDP.

Illegal activities

Again, as part of addressing a GNI ESA 1995 reservation, improvements were implemented to the measurement of illegal activities in Blue Book 2014, as described in the article Inclusion of illegal drugs and prostitution in the UK National Accounts. As a further refinement, allowance has been made to reduce the estimates made for prostitution in household final consumption expenditure by a factor to reflect the holiday taken by prostitutes. The Dutch have estimated the total holidays to be 12 weeks per year, and we have applied the same factor to the UK estimates.

Correction to gross fixed capital formation for improvements made to dwellings

During quality assurance of the gross fixed capital formation system, an error in processing was identified. An adjustment factor was incorrectly applied to the VAT rate in all years, and this is now being corrected at the first open period opportunity.

Correction to gross fixed capital formation for agricultural data

Discussions with the Department for Environment, Food and Rural Affairs (Defra) have identified a processing error in the calculation of gross fixed capital formation for the agricultural sector. Disposals were not being deducted correctly, leaving a net figure which was being artificially inflated. This has now been corrected at the first open period opportunity.

Improvements to own account construction

During Blue Book 2014, a GNI ESA 1995 reservation for own account construction (within gross fixed capital formation) was addressed as described in the article Changes stemming from improved comparability of Gross National Income measurement. At the time of implementation, the data up to 2010 in the reservation used a benchmark for self builds data which had a base year of 2006. AMA Research Ltd has now provided us with information which allows us to produce a more timely annual benchmark figure for self-build homes for all years from 2007 to 2013.

Transport for London capital stock changes

As part of Blue Book 2015 changes, some of the subsidiaries of Transport for London previously recorded as public non-financial corporations were reclassified as local government bodies. This was explained in the article Classification changes to National Accounts for Blue Book 2015. To complete the required actions within National Accounts, adjustments have now been made to the capital stocks series back to 2008 to incorporate this reclassification.

Natural gas imports from Norway

HM Revenue and Customs (HMRC) have amended the method for collecting data for the compilation of natural gas traded with non-EU partners. This change only affects non-EU imports of natural gas and has already been applied in the UK Trade statistical release for 2014 onwards. At Blue Book 2016 this change is being taken back to 2011 and so, for the purposes of the period being presented in this article, only 2011 is impacted.

Further detail for the above changes can be found in Impact on GDP Current Price annual estimates 1997 to 2011.

Open period revisions from 2011 to 2013 (Layer 2)

Every year the UK National Accounts sets an open period for the latest few years into which any agreed revisions to data and/or methods can feed in. For the 2016 accounts, the open period was set to 2012 to 2014. Supply and use balanced estimates for 2014 were produced for the first time in the 2016 supply and use tables (SUTs), but 2012 to 2013 estimates were published in the 2015 SUTs so can be compared in detail.

The revisions in this period reflect both the continuation of the methods changes introduced from 1997 to 2011 described in Layer 1, plus any new or revised production, income and expenditure source data. Notwithstanding the methods changes, revised data were provided for various of the GDP components, including:

• Output and intermediate consumption, via revised 2013 estimates from the Annual Business Survey and the PRODCOM survey (The PRODCOM survey collects the value and volumes of manufacturers’ product sales in the UK. It provides estimates of manufacturers’ sales by product, from businesses based in the UK.)

• Profits, via the new quarterly private non financial corporations (PNFC) data provided by HMRC.

Tables 3a and 3b show the revisions to GVA and GDP between the 2014 SUTs and the Layer 2 estimates. Tables A1.3a and A1.3b summarise the revisions to the main aggregates applied in Layer 2 from 2012 to 2013.

Revisions to 2014

The 2016 National Accounts include the first supply and use balanced estimates of current price GVA and GDP for 2014. They replace previously published figures for 2014 produced via the quarterly GDP estimation process. As such, revisions can only be derived at the main aggregate level and with the exclusion of intermediate consumption, which isn’t available from the short-run source data, therefore GVA is proxied by assuming that intermediate consumption changes at the same rate as output.

Tables 4a and 4b show the revisions to GVA and GDP between the last unbalanced set of Quarterly National Accounts estimates for 2014 (for reference period Quarter 4 (Oct to Dec) 2015, published March 2016) against those balanced via the supply and use process in 2016.

Measuring industries’ intermediate consumption by product

The UK supply and use tables (SUTs) make use of a vast array of information on production, income and expenditure, collected both by the Office for National Statistics (ONS) and other government departments, as well as other sources. With such a large number of data feeds, from time to time demands and priorities change the nature and timing of some of these sources. In the majority of instances, source changes can be dealt with by adjustments to production schedules and data treatments, but in a minority of cases, source data can become unusable or unavailable.

One of the main examples of missing data comes from the ONS Purchases Inquiry, which last collected data for reference year 2005, on an SIC 2003 and CPA 2002 basis. The results were used to provide detailed breakdowns of industries intermediate consumption by product, in conjunction with expenditure information gathered by the Annual Business Survey. Preliminary work to reinstate this complex survey began in 2012 and a major project is now underway to produce new results for use by 2017.

Implementation of the European System of Accounts 2010 (ESA 2010)

The 2014 UK National Accounts introduced the first of a wave of updates to bring them into line with the new ESA 2010 framework for measuring national accounts across the EU. The UK programme is set to continue until 2019, during which time significant extra resource is planned to implement the changes.

The single largest change to the SUTs required under ESA 2010 is that EU member states must provide balanced data both in current prices and in volume terms by 2018. This will require significant changes to the ONS’s existing compilation systems, as well as amendments to the fundamental approaches to defining and revising source information. A major work programme is already underway.