GDP Quarterly National Accounts: Quality and Methodology Information

Quarterly National Accounts: general methods

As part of the UK it is neither practical, nor conceptually possible, to compile a full set of national accounts for Scotland which are comparable with the range of content in the UK quarterly national accounts, quarterly sector accounts and quarterly balance of payments releases produced by the ONS. The components that we do produce are:

• monthly, quarterly and annual gross domestic product (GDP) estimates
• quarterly and annual household sector income and expenditure accounts
• Quarterly and annual public sector revenue statistics

As explained in this report, longer term trends are determined by the more comprehensive analysis that feeds through the UK Blue Book, UK regional accounts and Scottish Government Supply and Use Tables. The shorter-term estimates in the quarterly national accounts can therefore be understood to be relatively provisional in nature and designed primarily to meet the short term demand for timely estimates of the Scottish economy. At the same time, while some statistics can definitely be viewed as provisional, national accounts statistics are never final; they can always be revised due to such developments as improved estimation for earlier periods or the retrospective introduction of new definitions or classifications.

The most data-rich components of GDP are found in the output approach, where most industries are measured using detailed extracts of the same data sources, and with similar methodology, to the ONS statistics for the UK as a whole. Further information on these statistics is available in the methodology guide for GDP Output in real terms.

The data and methods for the other components of GDP can be broadly split into three main categories which will be referred to throughout the remainder of this report:

• Top-down estimates are based on taking a Scottish share of UK results using an appropriate data source for regional apportionment.
• Bottom-up estimates are made using data sources which relate directly to activity in Scotland, without requiring data for the comparable UK results. For example, bottom-up estimates are used for most of the GDP output approach by industry.
• Mixed top-down/bottom-up estimates include elements of both types, for example with different approaches used for sub-components which are combined together.

Quarterly Supply and Use Estimates

In common with the approach taken by the ONS for the UK as a whole, there are two main principles which underpin the ways in which GDP is estimated for Scotland in the quarterly national accounts:

1. In nominal terms, the level of GDP is best estimated using a supply and use table framework. The annual supply and use tables summarise all transactions for a detailed breakdown of goods and services between all sectors of the economy, including imports and exports, and account for total supply and total demand in the economy. In any given year, supply must be balanced with demand. These tables enable GDP to be presented for all three approaches using internally consistent values for the components, and provide a framework for making adjustments to estimates of the components to ensure that this is the case.
2. short-term growth in GDP, in both nominal and real terms, is best estimated using the output approach. The estimates for the other two approaches are therefore constrained to the output values. This principle applies to both the in-year quarterly and monthly paths during years which are covered by annual supply and use tables, and also to the provisional monthly, quarterly, and annual periods which are not covered by annual supply and use tables which is sometimes referred to as the short-term “tail” of the time series. 

The key way in which these principles are met is by the widespread use of methods to constrain quarterly estimates to the more comprehensive annual benchmarks provided by the supply and use tables. This is achieved using the Denton proportional difference benchmarking method described in the IMF Quarterly National Accounts Manual. Bringing these principles together, with quarterly series which are constrained to the annual tables, you can therefore think of the results in quarterly national accounts as being an extrapolation of the annual supply and use tables.

The approach used for the Scottish quarterly national accounts is novel, and takes the step of not only extrapolating each component of the supply and use tables but also bringing those series together into a series of quarterly supply and use tables (QSU) up to the latest quarter. This QSU system then provides the framework for estimating all current price estimates of GDP and its components in an internally consistent way, where supply and use of each product group are balanced in each quarter and where the components of GDP in each of the three approaches are equal, just like in the annual supply and use tables.

A strength of the QSU system is that it helps to confront the relative weaknesses in some data components. Whilst a methodology does exist to extrapolate all components of the supply and use tables, some of these rely on forecasts rather than data and have much higher levels of uncertainty than other components. The QSU system allows these estimates to be confronted for consistency each quarter and for adjustments to be made which improve the quality of the estimates. Ultimately, for each product group in the economy there is an adjustment made to the most uncertain component each quarter – usually rest of UK (RUK) imports or changes in inventories – which brings total supply and use into balance. In this approach there is no statistical discrepancy in the Scottish tables, i.e. no difference between the sum of the components of GDP measures by the income, expenditure or production approaches. These final residual balancing adjustments can be thought of as offsetting the sum total of any underlying statistical discrepancy plus the effects of alignment adjustments (timing differences within a year) and the residual error across all the components.

As noted above, although the QSU system is based on a supply and use framework, the estimates should be viewed as provisional, and are not based on the same depth of data content or detailed quality assurance as the annual supply and use tables.

Seasonal adjustment

The published time series for quarterly GDP and other statistics in the quarterly national accounts are seasonally adjusted. Seasonal adjustment is the process of modelling the regular movements over time which are linked to the time of year, and then removing those effects from the data. The adjusted series represents the underlying trend in the data, along with irregular (not seasonally recurring) features which are of interest. Non-seasonally adjusted versions of the data are also available in the supplementary tables.

 Three approaches to measuring GDP

Gross Domestic Product is an important measure of economic activity, and GDP growth is commonly used as the main indicator of economic performance for a country or area. There are three ways in which GDP can be defined and measured. In theory these approaches are just three ways of estimating the same thing, and  with a complete set of data they would all provide the same result.

• The output (or production) approach to measuring GDP estimates Gross Value Added (GVA) at a detailed industry level before weighting these together to produce an estimate for the whole economy. GDP is then calculated by adding the value of taxes (and subtracting subsidies) on products, such as VAT. Gross Value Added is the value of all goods and services produced less the value of goods and services used up in the production process. The output approach is used to produce estimates of growth in Scottish GDP in real (volume) terms and in current prices (nominal terms).
• The income approach to measuring GDP estimates income generated by production in the form of compensation of employees (income from employment) and gross operating surplus (profits, including self-employment income) for the whole economy, plus the value of taxes (less subsidies) on products and production. Unearned and redistributed incomes, such as interest, pensions, income taxes and social benefits, are not counted here. GDP measured using the income approach is only available in current prices (nominal terms) because there are no price measures for GOS and it cannot be deflated. 
• The expenditure approach to measuring GDP estimates all domestic final expenditure (consumer expenditure, government expenditure, and capital investment) within the economy, plus the value of exported goods and services, minus the value of imported goods and services. The components of the expenditure approach are widely used for economic analysis, modelling and forecasting of supply and demand in the economy. At the present time, estimates of GDP using the expenditure approach are available in current prices (nominal terms) as accredited national statistics, and in real (volume) terms as statistics in development.

Balanced GDP

Although GDP can be measured in three different ways, in theory each approach should give the same result. However, in reality the three approaches tend to produce different results because of factors such as sampling error or missing data. After a large amount of data for a complete year is available, the different sources of information are compiled into Supply and Use tables, and adjustments are made which produce a single set of numbers where all three approaches to GDP are balanced and consistent. In the quarterly national accounts, these adjustments are made at a higher level than for the detailed annual tables.

The following sections summarise the current data sources and methodology used for the income and expenditure approaches to GDP, followed by the other statistics in the quarterly national accounts. Further information on the output approach to GDP is available in a separate methodology guide.