What does the survey entail?
Purpose of the survey
To describe the structure of the Dutch agricultural sector (including data on holdings, livestock, crops, and other topics such as agricultural labour force and other gainful activities). The data are used for research and serve as a basis for political decision-making (at national and international level).
Agricultural holdings with an economic size of 3,000 Standard Output (SO) or more. SO is an economic measure for the size of an agricultural holding; it is based on the average annual yield per crop or animal category and is expressed in euros. Farms with an SO below 3,000 euros are very small, these might be farms with only 1 dairy cow or 1 acre of bell peppers, for instance.
Agricultural holdings. Agricultural holdings are defined as holdings where agricultural crops are cultivated or livestock are reared with the purpose of selling the resulting products.
As of 2016, registration in the Dutch Commercial Register (Handelsregister) with an Standard Industrial Classification (SBI) code for agriculture is leading in determining whether it concerns an agricultural holding according to the Agricultural census. This results in the closest possible adherence to the statistical regulations of Eurostat and the (Dutch) implementation of the concept of 'active farmer' from the Common Agricultural Policy (CAP).
Starting date/year of survey
Farm structure surveys have been conducted for over 100 years and held annually since World War II. Originally, the Agricultural census was a survey conducted by Statistics Netherlands (CBS); later on it became a joint survey conducted by CBS and the Dutch Ministry of Agriculture. As of 2002 it is part of a combined data collection (Gecombineerde Opgave) that is conducted by the Netherlands Enterprise Agency (RVO). This entails combining the collection of data for the Agricultural census with the acquisition of data for various administrative regulations.
The first provisional figures are published at the end of June. Revised provisional figures are published in September and November; the definitive figures in March of the following year. The publication of figures in June and September is done at the national level. Regional figures are published in November and in March.
During publication of the first provisional figures in June, not all declarations have been received and/or processed completely yet. At that time, only the most important plausibility checks have been carried out, and additional estimations are conducted for the non-response which are based on the declaration over the previous year. Data collection is finalised in September. This is when additional estimates take place as well as further analyses and plausibility checks.
How is the survey conducted?
Type of survey
The Dutch Agricultural census is an integral census. The questions in the survey for the Agricultural census are determined annually in a working group, in which almost all stakeholders are represented.
Data collection takes place via the CAWI method. The data collected are supplied to CBS by RVO.
The respondents are agricultural holdings.
Error detection and correction methods
Upon receipt by RVO, the data are checked at micro level for hard errors and corrected. During analyses by CBS, plausibility checks are conducted through relationship checks and year-on-year comparisons, among other methods. If necessary, enquiries are made at RVO, but never at the holdings themselves. Plausibility checks take place at macro and meso levels. Corrections take place at micro level.
Quality of the results
The census is comprehensive; the response rate is over 96 percent. Statistical reliability is therefore practically 100 percent. Reporting for the Agricultural Census is required by law. Since the data are used for the implementation of various administrative regulations (subsidies, Manure and Fertilisers Act), measurement errors are rare.
The figures are comparable over time. Regional figures are based on data from the municipalities where holdings are registered. Every year, there may be changes in municipalities and municipal boundaries. Municipalities may be dissolved or newly formed; boundary corrections also take place frequently. For regional figures, the situation as it was during the reporting year is taken as a basis.
In the case of substantial content changes, a number of years are recalculated (reversion). The final of year of reversion is then published according to both methods.
Description of quality strategy
See section on error detection and correction methods. Furthermore, to validate the figures they are confronted with data from other sources (e.g. derived surveys), trends and information obtained from external expert groups.
A quality report has been compiled in the framework of the Integrated Farm Statistics (formerly called Farm Structure Survey). This ‘European agricultural census’ is held three times over a period of ten years. The Farm Structure Survey was succeeded by the Integrated Farm Statistics (IFS) in 2020.
The National Methodology Report (NMR) is a quality report providing brief descriptions of the national history of the IFS and underlying (national) legislation, the main data groups, the completeness and the reference periods, as well as the accuracy and reliability. In addition, the report focuses on timeliness and accessibility, and presents an overview of the most important publications and trends. Finally, privacy and confidentiality regulations are discussed and an overview is presented of the statistical process with the main milestones and key figures.