The statistical process
Riding the wave of the digital revolution – starting in the 1960s – CBS began processing data electronically by means of computers. As computer performance grew rapidly, various databases could be linked to each other and standardisation took place, along with the emergence of the Internet. Dutch central government created a system of key registers based on digital data, collected by government organisations for administrative purposes.
The Statistics Netherlands Act of 1996 stipulated that CBS must reduce the administrative burden (from surveytaking) to a minimum and allowed CBS to access government key registers for statistical purposes. Wherever possible, CBS began a gradual process of switching from sample surveys among individuals and companies to data collection using administrative sources from other public authorities. Aside from government key registers, CBS has also begun using business records in recent years, e.g. scanner data from cash registers at supermarkets to calculate price developments, or data from energy companies on household and business energy consumption.
Furthermore, combining data from a variety of registers opens up new statistical capabilities. These capabilities increase exponentially with the growing number of data sources available to CBS. Hence, access to new data sources is essential in maintaining the quality of statistics. Over the past few years, more and more new data sources (including big data) have emerged, while methods of collection have become increasingly varied.
At the same time, the compilation of statistical information based on these new sources may pose challenges in conceptual and methodological terms. The question of what particular data should be collected is more and more a matter of technological advance. But most of all, modern developments unlock a myriad of opportunities. By combining all these data sources innovatively and reliably, greater possibilities are created to compile new, highly relevant, society-oriented and more detailed statistics, all the while keeping a sharp eye on privacy and reliability.
The statistical process – step by step
- The statistical process is conceived from a question. It may be based on a statutory national or European obligation or a request from a ministry, for instance.
- Then, the preparatory stage is entered. This stage consists of several steps: first, to identify what data are needed and what data are already available*; second, to determine whether any supplementary data are required and from which sources. Examples include registers, questionnaire surveys and big data. If existing sources are inadequate, a supplementary questionnaire survey may need to be prepared; this will entail a study of the sample design and sampling method, amongst other things.
- The third stage is data collection. This involves obtaining data from all the necessary sources and performing technical inspections. Where needed, survey questionnaires are distributed and response data will be collected and entered. Immediately upon receipt of these data, CBS removes all directly identifying personal characteristics. Please refer to our information on privacy in this respect.
- The fourth step comprises data analysis and processing, in which evident errors are identified and corrected. In addition, various methods are used to estimate missing data, based on other available information. This is added to the existing sources. If - in case of a survey questionnaire - a sample was drawn, the data are raised to match the requested population level.
- The fifth step concerns statistical security. Here, careful checks will be performed with respect to possible disclosure of individual data in cells which are intended for publication. There may be no risk that data of individual companies and/or persons can be derived from the data. If there is such risk, cells for publication will be automatically aggregated using software developed by CBS or values will be removed from those cells. CBS will never publish statistical information about individual persons and/or companies or provide this to third parties.
- The final step is publication. This important step entails dissemination of reliable and coherent statistical information among the largest possible audience. This is done via media or directly, through various channels (e.g. StatLine, website and social media) and in various forms (e.g. animations, press conferences, news releases, videos or reports). See also our publication policy.
*) The most important government registers which are used by CBS include:
The BAG (Key register Addresses and Buildings) houses the municipal basic data of all addresses and buildings inside Dutch municipalities.
The Netherlands’ Cadastre, Land Registry and Mapping Agency (Kadaster) is the source owner of the Cadastral key register (BRK).
The Personal Records Database (BRP) is the key register for personal data within the system of key registers.
- HR The Commercial Register (HR) is the base register containing details of all businesses and legal entities in the Netherlands.
The Key register WOZ (Valuation of Immovable Property) consists of various data which are needed to relate the value of immovable property to both the object and the owner/stakeholder. This includes the “assessed value” (WOZ value), a BAG-listed address, as well as a link to cadastral parcels and/or addresses and to BAG dwellings, plots, moorings and/or premises.