Statistical processing

Source data

Household consumption is primarily estimated based on indicators such as the retail trade index, price indices, and VAT data, which are always used as benchmarks for quarterly and preliminary (t 1) estimates. The Household Budget Survey (FU) is incorporated only in the preliminary national accounts (t 2). At the same time, industrial sales and purchases of goods and services are also used to reconcile product balances and functional distribution, covering approximately 300 products.

In the final annual national accounts (t 3), more time is devoted to data processing, and all sources are incorporated in a more complete and detailed form. At this stage, about 2.350 products are used, ensuring a more accurate and comprehensive estimation of household consumption as well as final quality assurance of the time series.

Frequency of data collection

There is ongoing collection of sources for households consumption. The frequency for each source depends both on how quickly the first version is used in the national accounts and on how quickly it is declared final.

Data collection

Most of the data for Households final consumption expenditure are collected via deliveries from the primary statistics in Statistics Denmark. In addition, supplementary information is collected to a lesser extent to ensure that the national accounts achieve full coverage, among other reasons. The supplementary information is primarily publicly available information on companies' activities, products, and development through information on companies' websites, in newspapers and the like, but can also be obtained through direct contact with the companies and authorities.

Data validation

In addition to the data validation performed by the primary statistics, a number of other data validation procedures are performed after the national accounts receive the data. Among other things, there are validation procedures that ensure internal and external consistency for certain variables. In addition, a number of confrontation routines are carried out to ensure that a correct financial interpretation of the (partial) results can be made. There is no uniform procedure for this, but time series are often used for the individual variables, where any breaks or other irregularities are detected. This may give rise to further studies of sub-areas where the producers of primary statistics are involved. The scope and depth of the different validation procedures depend on the time available for the calculations, which is considerably shorter for the quarterly and preliminary annual national accounts than for the final national accounts.

Data compilation

The data processing for household final consumption expenditure in the national accounts includes the integration, harmonization and validation of data from multiple sources.

Data from multiple sources are integrated through linking and harmonizing information from, among others, the retail trade index, the Household Budget Survey, foreign trade statistics, and public accounts. These sources are combined to ensure full coverage of household consumption of goods and services.

Information is grouped and classified in accordance with the concepts of the national accounts and the international consumption standard COICOP18. Sources are recoded to the national accounts’ units and chart of accounts so they can be consistently included in the supply-use tables.

Missing or incomplete values are imputed using relevant indicators, including developments in comparable sources and time series. Where direct data are unavailable, estimates are based on indicators such as retail turnover, results from the Household Budget Survey, and the preliminary national accounts.

Statistical units and variables are derived in accordance with the definitions of the national accounts. Household consumption is measured as the use of goods and services by households and is distributed across consumption groups and products.

No explicit weights are applied, but certain sources and estimates are incorporated with varying significance depending on coverage and quality.

Aggregation and calculation of summary data are performed by compiling checked and validated source data at the product level to totals for household consumption.

The results are compiled in the national accounts’ supply-use tables, where household consumption is a central component of use. After reconciliation of the tables, consumption is available in both current and constant prices.

Adjustment

Seasonal adjustment are carried out to account for normal seasonal fluctuations in the quarterly HFCE figures. The data is seasonally adjusted using the X-13 ARIMA software package with a fixed methodology. Models are evaluated and adjusted regularly, typically once a year. When performing seasonal adjustment, the effect of Easter is considered to some extent, and pre-adjustments are made if the effect is significant. No separate adjustments are made for working days or trading days. Experience shows that estimating the precise effect of Easter can be difficult, especially in years when Easter shifts between the first and second quarters compared with the previous year. Seasonal adjustments should therefore be interpreted with caution in certain cases.

Indirect seasonal adjustment are used, meaning that individual time series for different consumption categories are seasonally adjusted at a detailed level, and the seasonally adjusted totals for household consumption are formed by aggregating these detailed seasonally adjusted series. As a result, statistical discrepancies may occur between different methods of calculation, for example between purchases of goods and services that are not adjusted separately.