Statistical processing

Source data

It is a census where all companies producing compound feed stuff in Denmark delivers data to the statistics. At the moment 13 companies delivers data to the statistics. Until 2022, the part of the census that deals with the production of compound feed was a total count, while the composition of the compound feed was a sample and subsequent enumeration.

Frequency of data collection

Annual

Data collection

Data is collected electronically in a questionnaires, which the companies upload to Statistics Denmark. The statistics has an information page in Danish. The companies fill in and upload a spreadsheet, which consists of two sheets. In the first sheet, they indicate the quantities of the company's production of various compound feed, excluding exports, as well as their quantities of purchased supplementary mixtures, in own production, excl. export. In the second sheet, they indicate the quantities of a wide range of raw materials in their own production of compound feed, excluding export.

Data validation

Data from the individual reporting sheets is collected and partial quantities and total quantities are calculated and compared with previous periods. Data is first debugged and validated on a micro level, for example by comparing individual reports with reports from previous periods. If these comparisons give rise to questions, the company is contacted for an in-depth explanation and a possible re-report. Data is then checked at the macro level. This is done by comparing partial quantities, e.g. Ready-to-use feed mixtures for sows, and by comparing the total quantities, e.g. Cattle feed mixtures in total, with previous total quantities and the development in the cattle and pig populations.

Data compilation

The data processing is divided between Production of compound feed (PAF) and Content composition of the compound feed (FBL), as they end up in different tables, but the two data processing methods are identical. PAF is used both for the preparation of the Statbank tables FODER3 and FODER2, while FBL is only used for FODER2.

In 2022, the data collection and data processing for FBL was restructured so that the collection became identical to PAF. It also means a new data processing method for FBL, where the individual variables (e.g. wheat or fishmeal) are no longer calculated based on a sample and conversion factors, which were based on the quantities from PAF. For both parts of the count, the validated data is added together, and thereby the total production of compound feed and the total quantity of grain and feed materials used in the compound feed are obtained. The rearrangement of FBL results in a data breach in FODER2 from 2022, as there is no longer direct comparability in the use of the feed materials and hence the amounts of single feed substances, which are calculated on the basis of the amount in the compound feed. The old enumeration method has led to a certain overestimation of the use of some of the feed materials in feed mixtures. However, this cannot necessarily be seen in FODER2 in 2022, as the quantities of the individual feed materials used in feed mixtures fluctuate quite a bit from year to year. Data is not weighted or imputed.

Adjustment

Data is not corrected.