Estimation of value added and related economic indicators using small area estimates (SAE_WS)

This project aims to leverage small area estimation (SAE) techniques to develop estimates based on data collected through the Federal Statistical Office's (FSO) value-added statistical survey (WS). These estimates relate to elements such as gross value added (GVA). The ultimate goal is to enable analyses of growth and productivity trends. A central aspect of the project is to assess the effectiveness of using SAE techniques to construct estimates for areas of study that are not explicitly accounted for in the sample design.

Economic analyses rely on both structural data, such as employment figures, and also financial information about enterprises. Business is what drives a country's economic growth ‒ typically expressed as growth in GDP. The FSO produces a range of enterprise statistics that provide this kind of information. However, most of these statistics are not comprehensive and are based on sample surveys. The survey plan is generally established at a relatively aggregated level, such as by groups of economic activity, to ensure a certain level of statistical precision. This is also the case with the value-added statistical survey (Wertschöpfungsstatistik WS). 

The WS survey utilises accounting results to observe the structure and development of enterprise activities. It also provides the foundation for calculating the main macroeconomic aggregates in the national accounts. 

Key characteristics of the annual value-added statistical survey (WS): 

• The survey targets commercial companies in the secondary and tertiary sectors, excluding banks and insurance companies.
• The statistical population includes companies with at least 3 employees. 

The WS survey is thus an invaluable source of data for economic analyses. However, the survey has certain limitations in terms of the granularity of the results, such as at regional or activity level, as well as the ability to study sub-populations of companies that are not explicitly accounted for in the sample plan. 

This project seeks to explore new approaches to generate valuable data for economic analyses. In the context of the value-added statistical survey (WS), several sub-topics can be grouped under the ‘small area estimates based on the value-added statistics (SAE_WS)’ label. These include sub-populations of enterprises included in the enterprise groups statistics (STAGRE), regionalisation and granularity ‒ all topics which could be developed as sub-projects that aim to leverage the results of the WS survey.

The aim of the innovative Small Area Estimates based on the Value-Added Statistics (SAE_WS) project is to exploit the data from the value-added statistical survey in combination with other sources through modelling (prediction and imputation). Whether the focus is on analysing specific sub-populations of enterprises (i.e. group-controlled enterprises) or exploring finer levels of detail (e.g. using the NOGA), the project's goal is to ensure a level of granularity that goes beyond the methodological constraints of the original value added statistics survey design and stratification criteria. While the modelling approaches employed within the SAE_WS may vary depending on the desired outcome, all share the common objective of ensuring consistency with international standards, which serve as the reference framework for both the National Accounts (NA) and Structural Business Statistics (SBS). 

The project therefore encompasses a range of interconnected topics or sub-projects, each with its own distinct objectives, statistical procedures and desired outcomes. These features will be described in dedicated sub-pages. This approach aims to ensure a degree of flexibility in the way the work is carried out, and to allow the experimental results to be gradually incorporated into day-to-day statistical production in function of their maturity. 

In an initial phase, we examine the estimation of gross value added (GVA) in the context of the enterprise groups statistics (VAB-STAGRE). 

In the second phase, other areas of research could also be included in this project to exploit the potential of small area estimation (SAE) techniques. This could include areas such as estimation at more granular levels (e.g. 4-digit NOGA code, cantons and large regions, head offices of multinational companies or special purpose entities [SPEs]) which are currently not prioritised in the sample plan because of their impact on optimal sample allocation and therefore production costs. Other related areas could include extending the estimates to legal units with fewer than 3 jobs (cut-off sampling). The approach described for the VAB-STAGRE could be extended to other positions in the account sequence (especially property income).

Specific results are presented under each sub-project pages. The primary purpose of these pages is to illustrate the chosen approach for that particular area of research. The limitations of each approach will thus be described in the dedicated sub-pages. 

When interpreting the results published on this site, it is important to keep in mind that the statistical techniques used are still in an experimental phase and may be adapted over time.