Statistical processing
Contact info
Short Term Statistics, Business StatisticsKari AJ Arildsen
+45 3917 3609
Get as PDF
The survey is based on a sample of Danish retail trade enterprises. The sample includes approximately 2,200 enterprises, and at the time of the first publication, the figures for a month are based on responses from approximately 1.800 of these enterprises for the initial publication.
The sample consist of 42 subgroups and enterprises are sampled based on their share of the yearly turnover for the given subgroup. The companies are ranked from largest to smallest and the companies, whose rank constitutes the bottom 10 pct. of turnover for their subgroup when summed, are never selected to participate. The companies whose rank lies between 11 and 49 pct. of the subgroup’s yearly summed turnover, are randomly selected. Lastly, the larger firms whose turnover altogether lies in the top 50 pct. of the yearly turnover for their subgroup are always included in the sample. The companies are selected based on VAT-declarations to the Danish tax administration.
Source data
The survey is based on a sample of Danish retail trade enterprises. The sample includes approximately 2,200 enterprises, and at the time of the first publication, the figures for a month are based on responses from approximately 1,800 of these enterprises for the initial publication.
The sample consist of 42 subgroups and enterprises are sampled based on their share of the yearly turnover for the given subgroup. The companies are ranked from largest to smallest and the companies, whose rank constitutes the bottom 10 pct. of turnover for their subgroup when summed, are never selected to participate. The companies whose rank lies between 11 and 49 pct. of the subgroup’s yearly summed turnover, are randomly selected to participate. Lastly, the larger firms whose turnover altogether lies in the top 50 pct. of the yearly turnover for their subgroup are always included in the sample. The companies are selected based on VAT-declarations to the Danish tax administration.
Frequency of data collection
Data is collected monthly.
Data collection
The turnover figures are reported by the enterprises in the sample either via a dial-in phone solution or an online questionnaire. Completion of the questionnaires is mandatory.
Supermarkets, discount stores, department stores as well as service stations sell a very wide variety of goods. In order to be able to estimate the turnover index for three main commodity groups, the enterprises of these four industries are requested to break down their turnover in three main commodity groups: Food and other convenience goods, Clothing etc. and Other consumer goods. If the enterprise is unable to submit an accurate breakdown, it is requested to estimate the share of sales stemming from each group.
Online form and instructions can be found on the information page (only in Danish).
Data validation
The submitted data undergo troubleshooting. If the reported turnover deviates considerably from earlier records or from that of similar enterprises for the relevant period, then the enterprise in question is asked to investigate the correctness of the returned turnover figures. It is assumed that not all errors in submitted forms are detected, and therefore the statistics are subject to some uncertainty connected to reporting errors. The troubleshooting focuses on the reports with most impact on the results, and the uncertainty connected to reporting errors is therefore normally considered quite low, especially when it comes to the monthly growth rates on an aggregate level.
Data compilation
Totals are estimated from grossing up the reported data to the full population of retail trade enterprises, although enterprises below the cut-off limit are disregarded. For most industries, this limit is DKK 2.5 million in yearly turnover, and 1 million for a few industries, and the enterprises below the cut-off account for less than 5 per cent of the total retail trade turnover. In case of non-response amongst the largest enterprises, imputation is used. Usually, it is only necessary to impute values for less than 5 enterprises, corresponding to an imputation rate below 0.2 per cent.
In the estimation process, bias in the non-response pattern as well as in the sample is corrected by applying a ratio estimate, which includes information about the turnover subject to VAT for the previous year for the entire population of retail trade enterprises. This full population including information on the turnover subject to VAT is called the estimation population.
The calculation of the turnover totals behind the index figures is based on inter-monthly chaining. This means that every month growth rates are calculated by estimating totals for the month of interest and at the same time recalculating the previous month applying the same estimation population for both months. The ratio between these two totals is the growth rate for the month of interest. These growth rates are used to calculate turnover totals on which the index calculations are based, chaining on the actual estimated turnover totals for a single month. This month is currently January 2022.
The turnover is calculated by chaining on industry level. For the industries with turnover broken down by commodity groups, the chaining is performed on the level of commodity groups after which the industry totals are calculated by adding up the commodity group totals for each of these industries. The aggregate commodity group totals are calculated by adding up the relevant industry and commodity group totals.
Index values for a given month are calculated as the turnover of that month divided by the average monthly turnover of the base year (currently 2021). Consequently, turnovers are aggregated on industry level as well as commodity group and total level prior to the calculation of indices. In this way, no explicit weights are applied for calculation of aggregated indices. Instead, there is an implied use of internal weights, which are the turnover shares of the total turnover in the base year.
Only indices are published, not absolute turnover figures. For an idea of the level of absolute retail trade turnover of the individual industries, refer to the statistics on Purchases and sales by enterprises, which are based on VAT statements broken down by industries.
The volume indices for the commodity groups are calculated by deflating the value indices. This is done by dividing the turnover totals with appropriate price indices from the Consumer Price Index (CPI). The deflation is performed on the level of the commodity groups with the exception of the turnover of other consumer goods for the Service stations industry, which is deflated separately by means of a fuel price index. The deflated turnover totals are used for index calculations, and the result is a so-called indirect volume index.
Adjustment
Seasonal adjustment is applied to the commodity group figures as well as the Eurostat aggregates and this is done for value as well as volume indices. Indirect seasonal adjustment is used to calculate the aggregate seasonally adjusted figures, i.e. the seasonally adjusted figures for the commodity groups are calculated by means of weights from the base year. Each month, the figures are seasonally adjusted for the month in question.