Statistical processing

Source data

The Service Production Index statistics are based on existing statistics: Purchases and sales by enterprises. Producer Price Index for Services. Workplaces and jobs.

Frequency of data collection

Monthly and quarterly.

Data collection

The sources are the statistics covering Purchases and sales by enterprises, Producer Price Index for Services and Workplaces and jobs.

Data validation

Error detections are made at the industry and KAU level.

Data compilation

Indices on Production and Turnover in Service consists of a number of indices.

The turnover is delivered on legal unit. From the Business Register the is information on how many KAU, there is per legal unit. From the statistic Workplaces, job, full-time employment, wage and salary cost we get the average salary cost per workplace and their KAU. For every legal unit a total salary by KAU is calculated. The the turnover is split after share from each KAU in the legal unit.

The production indices are calculated as Index on Service Production = Index on observed monthly revenue (from Purchases and Sales by Firms) / Index on monthly Price (from Producer Price Index for Services). The indices from the Producer Price Index for Services are quarterly data. These quarterly data are converted to monthly data using the spline method.

Adjustment

To improve and harmonise the description of adjustment we answer the following questions:

Software used and version: JDemetra+, ver. 2.2.2 The model/filter selection (manual vs. automatic): Automatic How often are the models and the respective parameters re-estimated: Every three years The horizon of revisions (how often are the seasonally adjusted time series revised and how far backwards): From 2021, at every update Seasonal adjustment decomposition (additive vs. multiplicative): Multiplicative The model used: X13-ARIMA, [(2,1,0)(0,1,1)] The critical value for outlier detection: Automatically, determined by the number of observations in the (full) series span The filter length (automatically chosen vs. user-defined): Automatically The date of seasonal breaks in the series: None Indicate if indirect adjustment via components is used: Direct adjustment Indicate whether residual seasonality is checked and from which level of detail the aggregation is started: Not applicable Indicate the consistency amongst the different levels of breakdown: Not applicable

The index is produced as calendar adjusted (CA), seasonal adjusted (SA) and not adjusted. Further reading on seasonal adjustment.