National Accounts articles: Output approach to gross domestic product (GDP(O)) improvement report: June 2016

Since the September 2015 report, work has continued on the output approach to Gross Domestic Product (GDP(O)) Improvement Project. In March 2016, for example, we published two industry review reports: Water Transport (industry 50), and Postal and Courier Activities (industry 53). This report outlines the progress which has taken place since September 2015 and describes:

• GDP(O) Improvements for Blue Book 2016
• progress on completed and current industry reviews
• wider improvements including the VAT project

With each UK National Accounts Blue Book, revisions are caused by new data and data sources, improved statistical methods as part of the commitment to continuous improvements, and supply and use balancing.

For the years where National Accounts estimates are derived from supply and use balancing, revisions to GDP(O) are a consequence of the balancing process rather than the suite of short-term indicators which are the focus of this improvement programme. For Blue Book 2016, the years up to and including 2014 were supply and use balanced. During this process the expenditure approach to measuring GDP (GDP(E)) was deemed to provide a better indicator of the annual GDP path.

The impact from the short-term output indicators for Blue Book 2016, were due to the changes resulting from the water transport (industry group 50) industry review and the postal and courier activities (industry group 53) industry review.

2.1 - Update on GVA Weights, describing any significant changes

Each industry within GDP(O) has a weight representing its proportion of the sum of gross value added (GVA) produced by the economy in a given year. GVA is derived from outputs less inputs, or output less intermediate consumption. The weights used by GDP(O) are calculated as part of the supply and use balancing for each year from 1997 and revisions tend to be higher in the later years. In order for estimates to better reflect the changing industrial structures and prices in the economy, the last base year and reference year for chained volume estimates have been moved forward by one year, from 2012 to 2013, in Blue Book 2016 and Pink Book 2016. The weights for 2013 are also used in the later periods. Updated weights will decrease the importance of industries where GVA has fallen and increase the importance of industries where GVA has risen. This will necessarily produce the potential for revision to the overall GDP(O) index, although the growth rates of constituent industries remain unchanged, and their significance will vary each year. There were no significant changes to the GVA weights for 2013.

The Eurostat rating is guided by the Eurostat Handbook on Price and Volume Measures in National Accounts, 2016 and adopts a more pragmatic interpretation where this has proved necessary in setting guidelines for the more conceptually challenging industries. This rating of each industry has been re-assessed by our experts in line with new guidance which was published earlier this year. Each industry has been updated to reflect the change in GVA weight for Blue Book 2015.

The assessment for each industry can be found in annex 1 – the priority matrix 2016. It is important to note that the overall rating for the industry is based on the weakest lower-level component, and where the rating of the lower-level component exceeds a 10% significance threshold in terms of its contribution to the industry as a whole, it rates the industry as a whole as A, B or C on the basis of the quality of the weakest lower-level component. C rated industries are ‘in need of improvement’, B rated industries are deemed ‘acceptable’ and A rated industries are given the highest ‘appropriate’ rating.

Table 1 shows the 2016 Eurostat ratings compared with 2015, the change in the total rating is due to the change in GVA weight for 2013 and to changes in methods for Blue Book 2016.

Annex 2 lists the 12 industries classified as C for 2016, which remains unchanged from 2014 and 2015, and outlines the factors influencing that decision. The percentage of Blue Book accounted for by C-rated industries has increased due to the increased contribution of some of the 12 C-rated industries (mainly professional services) to Blue Book. There were 19 industries classified as C for Blue Book 2013. The changes to water transport in Blue Book 2016 mean that the rating for the water transport industry has improved from a B ‘acceptable’ rating to an A ‘appropriate’ rating.

3.1 – Completed Reviews

On 23rd March 2016 we published two industry reviews covering 1.05% of the economy:

• Industry 50 - Water Transport Industry Review

• Industry 53 - Postal and Courier Activities Industry Review

As part of the industry review process recommendations for improvement were also published in the reports. A full list of recommendations for the two reports can be found in annex 4.

3.2 - Current Reviews

Currently there are six reviews taking place covering 25.13% of the economy, electricity and gas (industry 35), construction (industries 41-43), publishing activities (industry 58), programming and broadcasting activities (industry 60), financial and insurance services (industries 64-66), professional services (industries 69-75) and social care activities (industries 87-88). The industry review programme proceeded more slowly than originally planned last year due to the priority given to the VAT Project.

In August 2016 we aim to publish a full review for professional services, alongside the initial reviews of industries 64-66 (financial and insurance services) and 87-88 (social care activities). In these initial reviews, we outline the current methodology and discuss current issues. In the second part of the reviews, we address the issues and provide recommendations for improvement. These will be published at a later date, winter 2016 for industries 87-88, and winter 2018 for industries 64-66.

69 – 75 – Professional Services

The main focus of the Professional Services review is on improving the deflator method. We are aiming to do this by replacing the currently used wages deflator with a more conceptually appropriate business prices deflator from the Service Producer Prices Index (SPPI). A review of the SPPI’s is taking place to ensure that the quality criteria for GDP(O) is met. The final report for this industry is due in August 2016

41-43 – Construction including 41.1 - Property Developers

The measurement of construction output has undergone significant change, with new interim price indices representing an important conceptual improvement. Our Construction Project team is working on a programme of continuous improvement of construction output, which will be completed in 2018, with deflator improvements to be completed and implemented for Blue Book 2017. Due to the large scale of this programme and recent changes, it has been decided to combine the industry review covering industry 41.1 (property developers) and the remainder of industries 41 to 43 (construction) into this programme. Therefore, the industry review will run alongside the overall construction project and the report is expected in 2018.

87-88 Social Care

The social care industry is challenging to measure due to the different methods of payment for care. Social care is largely funded fully or subsidised by government, with individuals making contributions to top-up care. However, it is difficult to meaningfully split the market and non-market aspects. The non-market element is estimated using public sector output, which is internationally recognised. This measures total inputs as a proxy for total outputs, using general government final consumption expenditure (GGFCE) data, which is sourced at current prices from the HM Treasury’s online system for central accounting and reporting (OSCAR) database on a UK basis. The market element is currently estimated using workforce jobs data.

Traditionally market output is measured using turnover from the Monthly Business Survey (MBS). However, this may not be the most effective method for this industry. It would require not only asking for business turnover, but also asking how much of that was from government, which would increase survey costs and the burden on business. It would also require a new sample for MBS as social care is currently not collected. To ensure the burden on business is not increased, alternative sources are being reviewed that could potentially have an impact on Household Final Consumption Expenditure (HHFCE).

There are also a number of other areas that need to be considered as part of the review. New legislation from spring 2016 gives more freedom of choice to recipients of social care. More direct payments will be issued which may impact how the industry will be measured. The review will also consider the measurement of non-core social care activities such as child care, pre-school care and the housing element of social care. Part one of the review, to be published in summer 2016, will cover the current methodology and describe the most important issues. Part two, to be published in winter 2016, will outline recommendations and address the issues raised in part one.

64 and 66 – Financial and Insurance Services

The initial review of financial and insurance services will be published in August 2016. This will describe current methods, summarise the significant change in methodology that has taken place since Blue Book 2007, quality assure administrative data and describe how the newly developed Financial Services Survey conducted by ONS could potentially improve methodology. The final report, in winter 2018, will also consider the impact of the Enhanced Financial Accounts project (previously referred to as the flow of funds project) alongside the data from the new Financial Services Survey and consider how volume estimates could be improved.

35 – Electricity, Gas, Steam and Air Conditioning Supply

The initial review of electricity, gas, steam and air conditioning supply will be published in winter 2016. This will describe current concepts, data sources and methods, evaluate them against international guidelines (primarily, the Eurostat Handbook on Price and Volume Measures in National Accounts), and identify areas for improvement. The final report, to be published in summer 2017, will consider how data sources and methods could be improved.

58 – Publishing Activities

The initial review of publishing activities will be published in winter 2016. This will describe current concepts, data sources and methods, evaluate them against international guidelines (primarily, the Eurostat Handbook on Price and Volume Measures in National Accounts), and identify areas for improvement. The final report, to be published in summer 2017, will consider how data sources and methods could be improved.

60 – Programming and Broadcasting Activities

The initial review of programming and broadcasting activities will be published in winter 2016. This will describe current concepts, data sources and methods, evaluate them against international guidelines (primarily, the Eurostat Handbook on Price and Volume Measures in National Accounts), and identify areas for improvement. The final report, to be published in summer 2017, will consider how data sources and methods could be improved.

3.3 - Priority matrix

A key feature of the industry review programme has been the use of a matrix to prioritise the industries in need of review. This has been driven primarily by the weight of each industry and, with the publication of Blue Book 2016, the matrix has now moved to using weights in proportion to gross value added for 2013 expressed as parts per thousand. The matrix has undergone some methodological changes over time but the new matrix for 2016 has been prepared using the same methodology as 2014. We will continue to produce the priority matrix to capture the progress being made. For details on how the matrix is produced please see the 2014 report.

4.1 - UK Statistics Authority Assessment and admin data recommendation

Following the UK Statistics Authority assessment of Short-Term Economic Output Indicators: Preliminary Estimate of GDP, Indices of Production and Services, and Retail Sales, in March 2015 it was announced that these 4 indicators would retain their National Statistics status. You can find further details in the confirmation letter.

The UK Statistics Authority made 13 recommendations with which the short-term economic output indicators must comply. One of the recommendations is to review the administrative data used to compile the short-term indicators. This recommendation is part of the UK Statistics Authority‘s Regulatory Standard for the Quality Assurance of Administrative Data, published in January 2015.

We have agreed with the UK Statistics Authority to conduct a review of all administrative data sources to be completed by 2018. This will continue to be part of the industry review process, where all data sources are reviewed for their quality and conceptual appropriateness. There are 4 industries, 01 (agriculture), 02 (forestry), 03 (fishing and aquaculture) and 94 (activities of membership organisations) that use admin data which will not have been reviewed by agreed deadline. A separate review of these 4 industries, to look at administrative data sources only, will be completed by 2018.

Details of the sources used to compile the short-term indicators can be found in the GDP(O) source catalogue, first published in 2014 and revised for 2015, and again for 2016. The catalogue of sources will continue to be updated on an annual basis, and published alongside Blue Book 2016 on 31 July 2016.

4.2 - Website update

As in previous years we are continuing to make improvements to the information available to users on the short-term indicators web pages. In 2014, we published the GDP(O) source catalogue for the first time, we published an updated version in 2015, and will be publishing the latest update alongside Blue Book in July 2016. Additional improvements have also been made, including the publication of our methodology for the UK SIC 2007 industry 01 (agriculture).

4.4 - Exploitation of HMRC VAT data

Over the last year significant progress has been achieved in exploring the use of administrative data as collected by Her Majesty’s Revenue and Customs (HMRC) from Value Added Tax (VAT) returns. To date the work has focused on the use of the HMRC turnover variable as collected on a VAT return with exploration as to its possible use within short-term output indicators and across National Accounts.

The project has established a monthly process that takes the administrative data, as collected at VAT registration unit level, links and cleans it to produce reporting unit level HMRC turnover data. This micro-data provides the basis for interrogation and analysis behind top level industry aggregates. This work has significant potential to save money by reducing ONS survey costs and the burden on businesses, in terms of completing government questionnaires. There are also expected to be quality gains arising from the increased sample size. To take this work forward, the proposed method is to combine administrative data with survey data and gain the advantages of both data sources. The proposed method which has been investigated over the past year is to use HMRC turnover data for smaller and medium sized businesses in conjunction with the Monthly Business Survey sample for larger businesses.

For further information on this work please see the following articles:

Feasibility study into the use of HMRC turnover data within Short-term Output Indicators and National Accounts (published 14 August 2015).

Exploitation of HMRC VAT Data (published 7 October 2015)

HMRC VAT Update December 2015 (published 21 December 2015)

HMRC VAT Update April 2016 (published 4 April 2016)

5.1 – PPI Rebasing

The Short Term Statistics (STS) Regulation requires all short term statistics to undergo price rebasing every five years, with the next rebasing from the current base year of 2010 = 100 to the new base year of 2015 = 100, required for delivery to Eurostat in 2018. Currently ONS is considering a number of options to meet this requirement and will provide further information in due course.

5.2 - Supply and use tables in previous year’s prices and deflator strategy

ONS has a long term aim to develop balanced supply and use tables, based on volume (previous year’s prices) in line with international best practice and is a requirement under European System of Accounts 2010 (ESA10). An important component of this process is the availability of appropriate output deflators that will also enable progress with “double deflation”. ONS is working to develop the deflator strategy across a variety of areas of the National Accounts. This will also support the methodology that will underpin the derivation of volume based tables for supply and use, particularly where the balanced volume estimates are fed back to the various component parts of the National Accounts. The first balanced supply and use tables in previous year’s prices will be produced to support Blue Book 2018.

5.3 – Framework Regulation Integrating Business Statistics (FRIBS)

The Framework Regulation Integrating Business Statistics (FRIBS) was included as part of the European Statistical System vision in 2014. It aims to apply consistent statistical definitions and ease the burden on business, for example through better use of administrative data. The regulation is still under discussion, but is expected that it will come into effect in 2019 at the earliest. The 2 most significant impacts for short-term indicators are to move the Service Producer Price Indices (SPPI) from “business to business” transactions to “‘business to all” and to redefine the retail trade.

ONS will also consider how consumer and export purchases should be included, as well as business and government purchases of services in a redefined SPPI. The new definition of the retail trade is expected to include the wholesale and motor trades industries. The most important impact is that for these 2 additional industries data would have to be produced significantly earlier than at present. These estimates will also have to be on “retail” or “trade” turnover and not the total turnover of businesses (so that non-retail turnover is excluded). This represents a significant challenge; the review to cover these industries has been timetabled for spring 2016 alongside the work to meet the new FRIBS requirements. The completed industry review is expected in winter 2018.

The next GDP(O) Improvement Report will be published in 2017 alongside Quarterly National Accounts consistent with Blue Book 2017. In August 2016 we will publish a final report into industries 69 to 75 (professional services) and 2 initial reports into industries 64 to 66 (financial service activities; insurance, reinsurance and pension funding; and activities auxiliary to financial services and insurance activities) and industries (87 to 88) social care.