Farm structure (ef)

Reference Metadata in Euro SDMX Metadata Structure (ESMS)

Compiling agency: Eurostat, the statistical office of the European Union

Eurostat metadata
Reference metadata
1. Contact
2. Metadata update
3. Statistical presentation
4. Unit of measure
5. Reference Period
6. Institutional Mandate
7. Confidentiality
8. Release policy
9. Frequency of dissemination
10. Accessibility and clarity
11. Quality management
12. Relevance
13. Accuracy
14. Timeliness and punctuality
15. Coherence and comparability
16. Cost and Burden
17. Data revision
18. Statistical processing
19. Comment
Related Metadata
Annexes (including footnotes)
National metadata

For any question on data and metadata, please contact: EUROPEAN STATISTICAL DATA SUPPORT


1. Contact Top
1.1. Contact organisation

Eurostat, the statistical office of the European Union

1.2. Contact organisation unit

E1: Agriculture and fisheries

1.5. Contact mail address

2920 Luxembourg LUXEMBOURG

2. Metadata update Top
2.1. Metadata last certified 06/03/2020
2.2. Metadata last posted 06/03/2020
2.3. Metadata last update 06/03/2020

3. Statistical presentation Top
3.1. Data description

The structure of agricultural holdings (collected through farm structure surveys - FSS) is presented at different geographical levels and over periods. The information follows up the changes in the agricultural sector and provides a basis for decision-making in the Common Agricultural Policy (CAP) and other European Union policies.
The survey unit is the agricultural holding (farm). The data on individual agricultural holdings are collected by all Member States, Iceland, Norway and Switzerland and sent to Eurostat. Montenegro, North Macedonia, Albania, Serbia and Kosovo* have also provided data for some years. The aggregated results are disseminated through statistical tables.
The FSS surveys are organised in all countries in a harmonised way. For a given survey year, countries have to conduct their surveys within the agreed time-frame. The characteristics are based on a legal act, the same data are available for all countries in case of each survey. Thus all the data are as comparable as possible.

*This designation is without prejudice to positions on status, and is in line with UNSCR 1244 and the ICJ Opinion on the Kosovo Declaration of Independence.

3.2. Classification system

Data are arranged in tables using many classifications. Some explanations are provided for most classifications.

The classifications of variables are available in Regulation (EC) No 1166/2008 (for the years 2010 and 2013) and in Commission Regulation (EU) No 715/2014 (for the year 2016). For the previous years, please consult List of Regulations and Decisions underlying the farm structure surveys in the annex of item 6.1.


Farm typology

The Community typology means a uniform classification of the holdings based on their type of farming and their economic size. Both are determined on the basis of the standard gross margin (SGM) (until 2007) or standard output (SO) (from 2010 onward) which is calculated for each crop and animal production. The farm type is determined by the relative contribution of the different productions to the total standard gross margin/standard output of the holding.

For more information on farm typology, see the Standard Output glossary article.

For the legal acts governing the typology system from 1990 onwards, please consult List of Regulations and Decisions underlying the farm structure surveys in the annex of item 6.1.

Territorial classification

The regional data is broken down according to the NUTS classification.

The regional data is available at NUTS level 2.

For the legal acts governing the NUTS classification, see item 6.1.

3.3. Coverage - sector

The FSS statistics cover agricultural holdings undertaking agricultural activities as indicated in item 3.5 below and meeting the minimum coverage requirements (thresholds) as indicated in item 3.6 below.

In countries where the survey threshold is above one hectare of utilised agricultural area (UAA):

3.4. Statistical concepts and definitions

Both the censuses and the sample surveys are aimed at producing a variety of information on specific CAP targets, as well as providing a basis for extrapolating Farm Accountancy Data Network (FADN) data. 

A set of characteristics and definitions are set in legislation, please consult List of Regulations and Decisions underlying the farm structure surveys in the annex of the item 6.1.  Some exceptions are nevertheless recorded. For 1990-2005, see annex  Exceptions and explanations on the characteristics of FSS surveys 1990-2005. Since 2007, see the national methodological reports

Information regarding the following groups of characteristics is available:

  • General information: location of the holding and farming system (organic farming, owner- or tenant farmed agricultural area).
  • Farm management: all persons responsible for the day-to-day management of the holding.
  • Farm labour force (Farm labour force - family labour, Farm labour force - non family labour, Farm labour force - not directly employed). Labour force is expressed in persons and in Annual Work Units (AWU) where one AWU corresponds to the work performed by one person occupied on a full-time basis.
  • Land area and its utilisation: size and distribution of the land area of the holding, in particular utilised agricultural area (UAA) which comprises arable land, permanent grassland, permanent crops and kitchen gardens.
  • Livestock: animals kept on the holding (cattle, sheep, goats, pigs, poultry, equidae and other animals expressed in heads or in livestock units (LSU) where 1 LSU is the grazing equivalent of one adult dairy cow).  
  • Machinery and equipments necessary to cultivate the land or manage the holding activity.
  • Secondary activities (activities directly related to the holding using the resources and/or products of the holding) and agro-environmental aspects.

A complete glossary of agricultural statistics is available for users.

Exceptions and explanations on the characteristics of FSS surveys 1990-2005
3.5. Statistical unit

The statistical unit is the agricultural holding.

3.6. Statistical population

The statistical population consists of agricultural holdings which meet the minimum requirements set up in the FSS legislative framework.

  • Until the FSS 2007, in accordance with Regulation (EEC) No 571/88, the countries which applied a threshold of above one hectare of utilised agricultural area (UAA) committed themselves to fix this threshold at a level excluding only the smallest holdings, which together contributed 1 % or less to the total standard gross margin (SGM).

  • From 2010 onwards, following the entry into force of Regulation (EC) No 1166/2008, the minimum requirements for survey coverage from the 2009/2010 FSS onwards have been modified, and countries which used a survey threshold above one hectare UAA were allowed to fix this threshold at a level that excludes only the smallest agricultural holdings which together contribute 2 % or less to the total UAA excluding common land and 2 % or less to the total number of farm livestock units (LSU). In all cases, all holdings above at least one threshold of a set of thresholds should be included. For this set of thresholds (set by legislation), as well as for the specific thresholds applied by countries, see Farm structure survey - survey coverage.
3.7. Reference area

The latest 2016 FSS covered the European Union, Iceland, Norway, Switzerland, Montenegro, North Macedonia and Serbia.

The 2013 data covered the European Union, Iceland, Norway, Switzerland and North Macedonia.

The 2010 census data covered the European Union, Iceland, Norway, Switzerland, Montenegro, Albania, Serbia and Kosovo*.

See for which country and year the data is available on Eurostat website, in the FSS glossary article.

*This designation is without prejudice to positions on status, and is in line with UNSCR 1244 and the ICJ Opinion on the Kosovo Declaration of Independence.

3.8. Coverage - Time

FSS has been carried out since 1966.

The online database is only available from 1990 onwards. See for which country and year the data is available on Eurostat website, in the FSS glossary article.

For earlier data (before 1990), see Farm Structure - Historical Results (surveys from 1966/67 to 1997).

3.9. Base period

The FSS data have been processed with SGM coefficients/SO coefficients (3-year average in the case of SGM and 5-year average in the case of SO). For more information, see the use of SO and SGM coefficients in Farm Structure Surveys in the period 1990-2013.

4. Unit of measure Top

Two kinds of units are generally used:

  • the units of measurement for the characteristics (area in hectares, livestock in (1000) heads or LSU (livestock units), labour force in persons or AWU (annual working units), standard output in Euro, etc.) and
  • the number of agricultural holdings having these characteristics.

5. Reference Period Top

FSS data are available for the following years: 1989/1990, 1993, 1995, 1997, 1999/2000, 2003, 2005, 2007, 2009/2010, 2013 and 2016. The agricultural censuses are in line with the FAO recommendations and are carried out every 10 years. The intermediate surveys are organised 3 (until 2007) or 2 times (since 2010) between the censuses. The exact reference periods are determined in legislation.

For 2010, 2013 and 2016, the reference periods are provided in Article 8 of the Regulation (EC) No 1166/2008. Article 16 of the same Regulation provides derogations for Greece, Spain and Portugal, where the references to the year 2010 are replaced by references to the year 2009. For the definition of the reference periods for groups of characteristics in legislation, as well as for the slight deviations from these reference periods in some countries in 2013 and their impact on the comparability across the countries and betweeen 2010 and 2013, see Farm structure survey - reference periods.

The surveys 1990 and 2000 covered a period between 1 December 1988 and 1 March 1991, respectively 1 December 1998 and 1 March 2001, thus the actual survey varied from country to country. See for which country and year the data is available on Eurostat website in the FSS glossary article.

6. Institutional Mandate Top
6.1. Institutional Mandate - legal acts and other agreements

The rules governing the farm structure surveys are laid down in a number of Council Regulations and Commission Regulations and Decisions, which are published in the Official Journal of the European Communities.

The annex List of the Regulations and Decisions underlying the FSS presents all legal acts by survey year and type of information provided (organisation of surveys, characteristics, definitions, tables, deadlines for data transmission, typology).

The NUTS classification is originally based on on the establishment of a common classification of territorial units for statistics. The Regulation also specifies stability of the classification for at least three years. Stability makes sure that data refers to the same regional unit for a certain period of time. This is crucial for statistics, in particular for time-series. However, sometimes national interests require changing the regional breakdown of a country. When this happens the country concerned informs the European Commission about the changes. The Commission in turn amends the classification at the end of period of stability according the rules of the NUTS Regulation. Therefore, the subsequent regular amendments to the annexes were adopted by Commission Regulation (EC) No 105/2007Commission Regulation (EU) No 31/2011, Commission Regulation (EU) No 1319/2013 and Commission Regulation (EU) 2016/2066 , establishing the NUTS versions 2007, 2010, 2013 and 2016.  In addition, the NUTS classification was completed with the regional breakdowns of the countries that joined the EU (please see Regulation (EC) No 1888/2005 and Regulation (EC) No 176/2008), and updated by  Commission Regulation (EU) No 868/2014, Commission Regulation (EU) No 1046/2012. For more information, see the NUTS website.  

List of the Regulations and Decisions underlying the FSS
6.2. Institutional Mandate - data sharing

Not applicable.

7. Confidentiality Top
7.1. Confidentiality - policy

Regulation (EC) No 223/2009 on European statistics,  amended by Regulation (EU) 2015/759 stipulates the need to establish common principles and guidelines ensuring the confidentiality of data used for the production of European statistics and the access to those confidential data with due account for technical developments and the requirements of users in a democratic society.

7.2. Confidentiality - data treatment

In the tables disseminated on Eurostat website, a cell is confidential if:

  • the (extrapolated) number of holdings that contribute to the cell (extrapolated) value is less than or equal to 4 and/or
  • one or two (extrapolated) holdings represent more than 85% of the cell (extrapolated) value.

A confidential value is replaced with ":c".

For non-confidential cells, the extrapolated number of holdings and all values of characteristics in cells are rounded to the closest multiple of 10.

Because of the confidentiality treatment, the sum of the individual cells does not systematically match with the value of the "total" cell.

8. Release policy Top
8.1. Release calendar

There is no release calendar. Only transmission deadlines are set in legislation.

8.2. Release calendar access

Not available.

8.3. Release policy - user access

In line with the Community legal framework and the European Statistics Code of Practice,  Eurostat disseminates European statistics on Eurostat's website (see principle 10 - 'Accessibility and clarity') respecting professional independence and in an objective, professional and transparent manner in which all users are treated equitably. The detailed arrangements are governed by the Eurostat protocol on impartial access to Eurostat data for users.

9. Frequency of dissemination Top

Results are disseminated 2 years after the reference year.

10. Accessibility and clarity Top
10.1. Dissemination format - News release

News release on Farm structure survey 2016

News release on Farm structure survey 2013

10.2. Dissemination format - Publications

The FSS results have been published in a series of statistical books, pocketbooks, Eurostat yearbooks and Statistics Explained articles, including a complete glossary:

Dedicated section:

Regional Data

Data for censuses carried out every 10 years are available in a three-level geographical breakdowns of the whole country, the regions and the districts; while data for intermediate sample-based surveys are only available upon the two-levels of country and regions.

Since FSS 1999/2000, information about local farm location is collected in most countries, so the data can also be disseminated by NUTS and are robust regarding the changes in the NUTS definition.

The FSS 2009/2010 information is in line with the NUTS 2010 classification: Commission Regulation (EU) No 31/2011 amending the NUTS classification from January 2003.

The FSS 2009/2010 and 2013 information are in line with the NUTS 2013 classification, see: Commission Regulation (EU) No 1319/2013 and Commission Regulation (EU) No 868/2014 amending the NUTS classification established by Regulation (EC) No 1059/2003.

The FSS 2016 information is in line with the NUTS 2013 classification.

FSS surveys do not cover the whole territory but only the land covered by the agricultural holdings. So the land use data without link with other farm characteristics should be downloaded by the user from the relevant domain.

10.3. Dissemination format - online database

The main dissemination outlet is the website of Eurostat. Aggregates at regional (NUTS 2), national and European level are published under node "Farm structure" (ef). 
The results span a considerable number of variables, broken down by groups of holdings formed according to several of the classification characteristics on which data were collected or compiled.

10.4. Dissemination format - microdata access

Eurostat grants access to microdata for scientific purposes only. Information on the procedure can be found in the dedicated page.  

Commission Regulation (EC) No 606/2008 included the FSS in the list of surveys that can be accessed for scientific purposes and the criteria used for the anonymisation are under preparation.

10.5. Dissemination format - other

Please consult free data online or refer to

10.6. Documentation on methodology
10.7. Quality management - documentation

11. Quality management Top
11.1. Quality assurance

Systematic data validation in countries and in Eurostat, see item 18.4.

National methodological reports, see items 10.6 and 10.7.

Peer review in Eurostat concluded in 2008.

Internal Audit in Eurostat in 2013.

11.2. Quality management - assessment

Data is collected from reliable sources applying high standards with regard to the methodology and ensuring a high degree of comparability.

The national methodological reports, sent by the countries for each survey, comprise information on each of the quality aspects defined by Eurostat. The national methodological reports are aligned to the European Standard Quality Reporting Structure.

12. Relevance Top
12.1. Relevance - User Needs

The users are all the stakeholders in the Common Agricultural Policy (CAP) and other policy areas such as environment, regional development, climate change and health:

  • policy makers: DG Agriculture, National Ministries of Agriculture, European Parliament etc.;
  • reviewers, assessers, analysts;
  • professional groups (unions, press);
  • researchers.

The list of tables and of the surveyed characteristics are approved by Comitology.

12.2. Relevance - User Satisfaction

No feedback analysis.

12.3. Completeness

According to the Regulation (EC) No 1166/2008, if a country determines that a characteristic has a low or zero prevalence, the characteristic may be excluded from the data collection. 
For each survey, the countries fill in a questionnaire by indicating the list of non-existent (NE) and non-significant (NS) characteristics with related reasons and the used sources. These lists can be accessed in the NMRs.
A NE characteristic has 0 value for the data, because it is for example not grown/raised due to climate conditions.
The value of a NS characteristic is not significant as compared to the whole population, because it is for example not grown/raised due to climate conditions or it is not profitable enough. A NS characteristic (with low prevalence and not collected under the heading) is denoted with "0n"  in the tables on Eurostat website.

For designating a characteristic non-significant, Eurostat recommends the following general thresholds:

  • area is less than 0.01% of total UAA;
  • the LSU is under 0.5-1% of total LSU;
  • the share of holdings with the characteristic in question (e.g. "more that 50% of the value of the final production self-consumed by the holder", "other gainful activities") is less than 2% of total number of holdings.

However, the designation of a characteristic as NS is subjective and it is not possible to give a threshold fully applicable to all countries and all characteristics. In other words, thresholds are different for different agricultural crops and livestock units in different countries.

13. Accuracy Top
13.1. Accuracy - overall

Regulation (EC) No 1166/2008  requires the estimates to be representative at the level of NUTS2 regions and by farm type and by size of agricultural holdings. This means that the estimates should be reliable and can be disseminated without quality concerns at NUTS2 regions. This does however not hold for lower geographical levels, e.g. NUTS 3, unless for countries which stratify by NUTS 3.

The sample surveys are usually stratified at NUTS 2 level.

The results of sample surveys are extrapolated with one (sometimes two) weighting factor(s) for a wide range of characteristics. Therefore the accuracy of estimates is affected by sampling errors.
In addition, accuracy is also affected by non-sampling errors which cause both variability and bias in the estimates.

13.2. Sampling error

Regulation (EC) No 1166/2008 provides for two levels of precision requirements, one for the sample (i.e. non-census) surveys and another for the SAPM in 2010. For the FSS, precision requirements for a list of 21 characteristics are defined, which apply to all NUTS 2 regions with at least 10 000 holdings. For a NUTS 2 region with fewer than 10 000 holdings, those precision requirements apply instead to the associated NUTS 1 region, provided that the associated NUTS 1 region contains at least 1 000 holdings. For the SAPM, a reduced list of 10 characteristics is defined. The precision requirements are applied only if the prevalence of the characteristic surpasses a certain threshold in the respective region.
Eurostat publishes estimates for many more characteristics than those for which precision requirements are set in the Regulation (EC) No 1166/2008 and for many types of population breakdowns, while the variables determining these breakdowns are not stratification variables in the sampling design.  Since some cells' estimates,  being of poor quality, likely send misleading messages to data users, the following quality rating system is implemented for cells' estimates:

  • for totals and means of continuous variables - the coefficient of variation (relative standard error) flag system:
      -below 24.99% - estimates are released;
      -25.0% - 34.99% - estimates are released with warning, to be used with caution; released cell values are followed by letter 'u';
      -35.0% and more - too unreliable, estimates are not released;  cells are flagged with 'u'.
  • for count indicators and proportions - the standard error flag system:
      -below 12.49 percentage points - estimates are released;
      -12.5 percentage points - 17.49 percentage points - estimates are released with warning, to be used with caution; released cell values are followed by letter 'u';
      -17.5 percentage points and more - too unreliable, estimates are not released;  cells are flagged with 'u'.

The national methodological reports provide the relative standard errors (RSEs) for the main crop and livestock characteristics, for which precision requirements are set in legislation.  The NMRs also provide details on the applicability of precision requirements and the methods used by countries for the estimation of the RSEs.  Finally, the NMRs indicate the reasons for possible cases where precision requirements are applicable and estimated RSEs are above the thresholds (the reason is usually the variability of the characteristics between the sampling design and the data collection).

13.3. Non-sampling error

The national methodological reports provide all reasons and treatments on the non-sampling errors including coverage errors, measurement errors, non-response errors and processing errors.

Non-response errors are generally very low because the legislation in place obliges response with fines in many countries in case of non-response.

Processing errors are also low because the on-line surveys and questionnaires embedded in IT applications include consistency checks and validation rules.

Seasonal adjustment is not applicable. On the other hand, a few countries apply calibration but the related error is not estimated.

For the FSS characteristics with the largest measurement errors, see the article Farm structure survey - measurement errors

14. Timeliness and punctuality Top
14.1. Timeliness

Deadlines for transmission of FSS data are set by legislation; for the 2010, 2013 and 2016 surveys in Regulation (EC) No 1166/2008 (for earlier surveys see item 6.1).

For the FSS 2013 and 2016, countries shall transmit validated survey data to the Commission within 12 months of the end of the survey year. Data relating to the rural development measures may be transmitted to the Commission separately within 18 months of the end of the survey year.

Results are disseminated following validation, confidentiality treatment and tabulation 2 years after the reference year.

14.2. Punctuality

No delays after validation of the data.

15. Coherence and comparability Top
15.1. Comparability - geographical

Comparability across countries is considered high. Harmonisation of concepts and methods is ensured by a number of Commission Regulations and Decisions providing detailed information on list of variables/characteristics, on rules and procedures that should be followed and on time-frame within which FSS should be conducted. The same data are available for all countries. Besides, the censuses carried out every 10 years are in line with the FAO recommendations enhancing this way the comparability with countries outside Europe including USA and Japan. 

Comparability problems between European countries concern:

  • Survey coverage
    Each country is allowed, in line with the legislation, to set up thresholds at a level that excludes very small holdings, as long as the conditions for minimum coverage are guaranteed. Consequently, all countries have set up a system of country-specific alternative thresholds, which ensure that the requested minimum coverage is respected. For more information, see items 3.3 and 3.6. For the period until FSS 2013 see the article Farm structure survey - survey coverage. For 2016, see the national methodological reports.

  • Geographical location of the holding
    For 2009/2010 and 2013, the methodology for geo-referencing the agricultural holdings is based on the geographical coordinates. Regulation (EC) No 1166/2008 defines the "location of the holding" as the latitude and longitude coordinates within an arc of 5 minutes that avoid the direct identification of an individual holding. If a latitude and longitude location contains only one agricultural holding, then this holding shall be attributed to a neighbouring location, which contains at least one other agricultural holding.
    In each country the geographical coordinates were attributed to each holding using distinct methodologies. See national methodological reports.
    For 2016 the transmission of the latitude and longitude coordinates to Eurostat was no longer mandatory. Data include only the NUTS 3 regions where the holdings are located, as according to Commission Regulation (EU) No 715/2014.

  • Regional breakdowns
    The territorial classification of regional data is broken down according to the NUTS classification.
    The regional data is available at NUTS level 2.
    FSS 2016 data use NUTS classification 2013; FSS 2009/2010 and 2013 data use NUTS classification 2013; and the 2007 and 2005 data use NUTS classification 2010. 2007 data use NUTS classification 2006.

  • Common land
    The way common land has been included in the survey has an impact on the comparability of number of holdings, average size of holdings and livestock densities across countries and by dimensions within each country.
    In countries where all common land can be clearly rented by or allotted to an agricultural holding (based on written or oral agreement) (Method A), it is not considered any more common land in statistics, but land normally used by the agricultural holding. In countries where common land is neither rented by, nor allotted to the agricultural holding (Method B), this land is actual common land in statistics and 3 different options (or some combination of these) to record common land areas have been used. For the presentation of the methods and options used by countries and the comparability problems for the period until 2013, see article Farm structure survey - common land. For 2016, see the national methodological reports.

  • Annual working unit (AWU)
    For the definition and comparability of the annual working unit (AWU) across countries, see article Farm structure survey - measurement errors.

  • Irrigation
    The area of crops which have actually been irrigated at least once during the 12 months prior to the reference day of the survey do not include crops under glass or other (accessible) protective cover and kitchen gardens which are almost always irrigated. If more than one crop is grown in a field during the harvest year, the area should only be indicated once: for the main crop, if irrigation was used for it, or otherwise for the most important irrigated secondary or successive crop.
    In FSS 2010, the volume of water that has been used for irriga­tion on the holding during the 12 months prior to the reference date of the survey, regardless of the source, is provided using data estimation, imputation, or modelling methods. This might affect comparability across countries.

See in addition the annex Exceptions and explanations on the characteristics of FSS surveys 1990 - 2005 in item 3.4.

15.2. Comparability - over time

The time series analysis in agricultural domain is mostly oriented to economic and social aspects of the agriculture such as the evolution of the labour force, the level of investment on the farms regarding environmental issues and the most relevant type of tenancy on the agricultural sector.

Real changes in the evolution of agricultural holdings

There is a decreasing trend in the number of agricultural holdings that appears in general over all European countries. Among the most important reasons to explain the changes over time are the abandonment of the agricultural activity, the lack of profitability of some crops and the increase of energy crops, based on the demand and subsidies. A time series analysis over a three year period helps to understand the impact of the agricultural policy on the agricultural practices i.e. the discontinuation of subsidies can lead to a quick and drastic decrease in the area used for the cultivation of a specific crop.

Methodological reasons affecting comparability over time

Changes are in many cases explained by the modification of the survey coverage (mainly thresholds, but also geographical coverage and coverage of farm types) and by treatment of common land.

  • Change of survey coverage

    From FSS 2007 onwards "maintaining land in good agricultural and environmental conditions" became an agricultural activity.

    The change of legal requirements on survey coverage (see item 3.6) led to changes of survey thresholds in some countries between 2007 and 2010 FSS.

    In the FSS 2013, an additional adjustment of thresholds took place in some countries (Belgium, France, Croatia, Italy, Luxembourg, Malta, Poland, Finland and Iceland).

    For more information on comparability problems and the impact of survey coverage changes on national data until FSS 2013, see article Farm structure survey - survey coverage.

    Information on the significant changes on survey coverage between FSS 2013 and FSS 2016 is presented below.

    • Italy changed the thresholds for agricultural holdings between the 2013 and 2016 surveys. This change had a significant impact on number of holdings and labour force directly employed by the holdings and did not have a significant impact on utilised agricultural area, livestock and standard output:
    • While the dataset shows that the number of holdings increased by 13.4% between 2013 and 2016, in reality, when comparing over the same coverage, the number of holdings increased by only 1.8% between the two years.    
    • While apparently the labour force directly employed by the holdings (in annual working units) increased by 5.2%, it actually increased by only 2.6% between 2013 and 2016.
    • The impact of change of thresholds is minor for utilised agricultural area  (an apparent increase by 4.1% compared to a real increase by 3.5%) and for total standard output (an apparent increase by 18% compared to a real increase by 17.8%).
    • For livestock units there is no impact (increase by 1.0% in both cases).

    • Malta: The number of holdings in 2016 and 2013   is underestimated as in these two years  the surveys did not include a part of holdings i.e. mainly those  which in 2010 only maintained land in good agricultural and environmental conditions. The comparability over time is not affected for most indicators; exceptions are number of holdings, fallow land and kitchen gardens.

    • Netherlands: Between 2013 and 2016, the statistical office changed the registers used to identify the agricultural holdings. The change of registers has almost no effect on arable land, permanent crops and most livestock indicators. It has a high impact on sheep, equidae and "other land" and a limited impact on goats and cattle. The decrease of sheep is caused by the exclusion in 2016 of holdings which keep sheep mostly for nature preservation or for hobby farming and not sheep for meat or milk.

  • Change of method or methodological problems on the collection of common land data
    Common land has not been consistently covered by all countries in FSS before 2010, despite that it has been covered by legislation at least since 1988. Regulation (EC) No 1166/2008 made it clearer that the data on the UAA should cover common land in all countries. Therefore, since 2010 FSS, common land has been included in the surveys by countries in a more harmonised way. 

    Nonetheless, over the years, some countries changed the methodology of collection of information on common land or faced problems on the identification of common land units.

    For more information on methods and impact of change of methods on the data until FSS 2013, see article Farm structure survey - common land.

    Between FSS 2013 and FSS 2016, the methodology changed at the level of the EU: in FSS 2016 the common land was newly recorded  under the new tenancy type "common land".  This triggered some fluctuations in the utilised agricultural area classified in the various classes of tenure. In addition:
    • Cyprus made a change in the methods used to collect data on common land in FSS 2016 and as consequence revised the data for 2013.
    • Croatia also made a change in the methods, however mentioned that the decrease of common land between 2016 and 2013 is caused by the fact that a part of common land is purchased or rented out and is no longer common land.
    • Italy newly considered in FSS 2016 the option of a municipality or institution managing its own land.
    • In FSS 2016, Poland newly collected data on common land (from administrative sources) in the form of common land units. Common land was no longer added to the leased land but recorded separately.  To avoid that farmers keep reporting common land under leased land, during the training of interviewers as well as in the methodological manual it was underlined that the area of the agricultural land does not contain common land.
      For more information, see the national methodological reports 2016.


  • Change of the definition of agricultural holding
    The modifications undertaken by countries on the definition of agricultural holding were mostly minor and had only a negligible impact on the number of holdings.
    For the analysis of changes between FSS 2010 and FSS 2013, see article Farm structure survey - definition of agricultural holding.
    Concerning the changes between FSS 2013 and FSS 2016, Germany newly covered holdings with more than 1000 places for poultry, even if they have no agricultural activities (including no poultry breeding) in the reference period.

  • Change of the reference period
    Some countries changed the reference periods within the provisions of Regulation (EC) No 1166/2008.
    The article Farm structure survey - reference periods presents the changes between 2010 and 2013 in the countries where they occurred. The modifications undertaken by countries on the reference periods/days in FSS 2013 as compared to 2010 were mostly minor and most probably had only a negligible impact on the number of holdings. However, for specific characteristics such as pigs, the modification of the reference period/day, for example from 1 December 2010 to 31 December 2013 has some effect on data comparability over time.
    Concerning the changes between FSS 2013 and FSS 2016, the following is noted:
    • In Belgium, the reference day for livestock characteristics changed from 15 May to 15 October, while the reference period for labour force characteristics changed from 1 May - 30 April to 1 October – 30 September.
    • In Bulgaria, the reference day for livestock characteristics changed from 1 November to 1 September, while the reference period for labour force characteristics changed from 1 October - 30 September  to  1 September – 31 August.
    • In Spain, the rural development measures concern 2 years in FSS 2016 (instead of 3 years in FSS 2013). The new Common Agricultural Policy entered into force late and some countries did not have time to implement the system in 2014.
    • In Montenegro, the reference day for livestock characteristics changed from 31 May to 1 November, while the reference period for land and labour force characteristics changed from 1 June - 31 May to 1 November – 31 October.
    • In Portugal, the reference day for livestock characteristics is a fixed date (compared to FSS 2013 when it was the day of the interview).


  • Change in the list of characteristics
    See Annex List of Regulations and Decisions underlying the farm structure surveys in item 6.1 for the Regulations on the lists of characteristics in various survey years.

  • Changes in the concepts and classifications
    - FSS 1999/2000 introduced a new concept of "group holdings" with an impact on the classification between family and non-family labour force.
    - From FSS 2009/2010 onwards, the classification of fruit and berry plantations changed from

- Fruit and berry plantations temperate climate

- Fruit and berry plantations subtropical climate


- Fruit species temperate climate

- Fruit species sub-tropical climate

- Berry species.

  • Changes in the geographical breakdowns
    Up to FSS 1997, the geographical breakdowns of the countries are based on specific regions and districts.
    From FSS 1999/2000 onwards, the data are displayed following the NUTS.
    FSS 2016 data use NUTS classification 2013. FSS 2009/2010 and 2013 data use NUTS classification 2013; and the 2007 and 2005 data use NUTS classification 2010. 2007 data use NUTS classification 2006.


15.3. Coherence - cross domain

Eurostat undertakes cross-check analyses of the FSS statistics against animal and crop (ACS) production statistics but also against Agricultural Labour Input variables included in the Economic Agricultural Accounts (EAA). The main reasons for the differences between FSS and other data collections are:

  • Different survey coverage
    Countries are using different surveys as sources for the data sets because they have different aims. One is a structural survey which is intended to gather information on the farm size and main land use as well as all the factors involved at farm level while the production surveys reflect multiple cropping areas and actual production during the reference year.  A main reason for the differences in the results is the different extent of survey coverage explained mostly by the different thresholds used by countries in different surveys. Full consistency between the surveys is not expected, but there is an aim to improve coherence between the figures of the agricultural statistical domains.

  • Different sources
    The main sources used are territorial surveys, administrative sources and a combination of them with diverse methods of grossing up. The use of administrative sources is an important factor of differences for pulses, fallow land and permanent crops. Better integration between administrative sources and surveys is sought.

  • Variability of the estimates produced from sample surveys
    Crops which have a very small cultivation area and higher sampling errors tend to show larger discrepancies. This happens particularly in the case of flowers and ornamental plants, other arable crops, some cereals and sunflowers. Industrial crops are also affected, for example, the area for tobacco, hemp, flax, aromatic plants, etc. Amongst animals, bovine over two years old (males and heifers) are affected.

  • Different concepts and definitions

Land use
The FSS does not cover the whole territory but only land belonging to agricultural holdings. The total farm area covers utilised agricultural area (arable land, kitchen gardens, permanent grassland and permanent crops), area covered by mushrooms, unutilised agricultural area, wooded area and other land (occupied by buildings, farmyards, tracks, ponds, rock, etc).

Utilised agricultural area
Some discrepancies in the utilised agricultural area are mainly due to the treatment of common land. The ownership/management of farms by public entities should be included under the scope of both FSS and crop production surveys to reduce discrepancies.
On arable land, differences are detected also due to other minor reasons such as pulses/fodder classification, fallow land concept or other items that could be grouped in different ways.
Arable crops and vegetables of the ACS use sown/harvested area which in most cases is bigger than the main area in the FSS.  In crop production surveys, the figures for area refer to area under cultivation/ harvested area whereas in FSS the area shall be counted only once (main area). This means that in FSS, although the area is used more than once in a certain calendar year it is counted only for the main crop on that area. So, in case of combined or successive cropping the area is counted in a different way than in ACS. The main crop is the crop with the highest value of the production or - if that is not determining enough - which occupies the ground for the longest time. 
The overall variability between planted and harvested area is not well known, usually less than 20% for cereals in Europe, but depends on several factors such as the method of preparation, the time and manner of sowing, the kind of land on which the crop is grown, and the climatic conditions of the particular section. Also, the area sown or planted for the harvest of a given crop may change throughout the growing season. Issues such as use for purposes other than grain, abandonment, extreme weather damage, or unusual economic conditions might also create a problem for the areas recorded in the different surveys.
Cultivated area in FSS for vegetables, melons and strawberries must be recorded only once even in the case of successive harvests whereas the area in crop production statistics is recorded as many times as it is cropped. This can be an important reason for different results especially for vegetables.
As regards permanent crops, the difference between intensive and extensive fruit farms is explained by planting density, fertiliser application, irrigation for higher production per area, etc. The FSS survey includes young fruit and berry plantations which are not yet in production, abandoned permanent crops with a possibility of reversibility and isolated trees. However, all these areas are excluded in the crop production survey and this is another source of the discrepancies in some countries.
As regards organic farming, Eurostat collects data on organic farming from national Certification and Inspection bodies. This data collection can differ from the FSS data on organic farming due to the different statistical unit (for organic statistics the unit is the certified agricultural holding) and to the data collection methodology (organic farming data is taken from administrative registers).
In order to increase coherence between FSS and ACS, the FSS 2016 Handbook was redrafted to harmonise as much as possible with the definitions and notes of the ACS Handbook, what is more, the proposal for the list of variables for 2020 onwards has been further harmonised with the ACS crop lists.

It is known that seasonal effects can appear for all species and in every production stage. Due to these effects the total number of heads and the production level recorded changes based on the different months of the year.
The bovine animals usually operate on a fairly stable cycle between the climatic seasons, unlike swine, sheep and goats. However, during the survey, farmers tend to supply late responses and to estimate the number of animals that they had at the time. Therefore, in relation to "2 years old heifers" and "other cows, bovine 2 years and older", several inconsistencies appear to be based on misclassification due to shortcomings on providing accurate replies regarding the age at disaggregated level.
For pigs, seasonality factors such as heat stress can affect the litter size and thus the production flow in a later production phase. Understanding seasonal effects helps to follow-up the changes in the number of heads and consequently a higher meat production level during some periods of the year.
The enumeration of sheep and goats is one of the most affected by seasonality. Usually the more lambs that can be born into a flock, the more profitable the operation is. This means mainly that maintaining ewes can be supported by the sale of many lambs. Obviously, this affects the inconsistencies on the total number of sheep and other sheep figures produced in summer and at the end of the year. Additionally, these variations are linked to the slaughtering and the big consumption of these species according to the market demand.
The seasonally different figures show that sheep, goats and pigs had maximum slaughtering in December and January indicating high demand for meat in these months while they declined to the lowest levels in August and September.

15.4. Coherence - internal

The results of the structure surveys of agricultural holdings are disseminated through statistical tables on Eurostat website. A wide range of tables are available crossing various dimensions and variables of time, area size, economic size, or geographical breakdowns. All the tables are derived from the same sets of individual data, which guarantees their coherence.

The internal coherence of the data sets is guaranteed by a common set of validation rules for the individual data and stable definitions of the variables/dimensions.

16. Cost and Burden Top

Some coordination exists between surveys in some countries to avoid that farms need to answer multiple questionnaires with the same kind of questions.

More information for each country can be found in the national methodological reports.

Legislation is in place to provide financial support for national FSS surveys.  

17. Data revision Top
17.1. Data revision - policy

Please refer to Eurostat policy on the management of errors in disseminated data.

17.2. Data revision - practice

The individual data are validated by countries and Eurostat using strict rules, the aggregated data are checked as well.

The 2013 and 2016 data were disseminated as preliminary and then as final.

These data may nevertheless be subject to subsequent revisions.

Regional data

In ef_main tables, the 2007 data were revised to set the NUTS classification from NUTS 2006 version to NUTS 2010 version.

18. Statistical processing Top
18.1. Source data

Countries collect data through two kinds of surveys:

  • an agricultural census (AC) every 10 years,
  • several sample-based surveys between them.

However, for certain characteristics, the countries may use samples for every survey.

The sampling rate depends upon the country and the survey year. It varies between 2.5% and 100% of the total population of agricultural holdings. In a few countries, every survey is a census, see the surveys between 1999/2000 and 2016 in the Glossary: Farm structure survey.

Countries complement the data from surveys with data from administrative sources.   The use of administrative sources is for the first time enshrined in Regulation (EC) No 1166/2008. According to Article 4 of Regulation (EC) No 1166/2008, provided that the information from the administrative source is of at least equal quality to information obtained from statistical surveys, countries are allowed to use information from:

  • The Integrated Administration and Control System (IACS) provided for in Council Regulation (EC) No 1782/2003,
  • The System for the Identification and Registration of Bovine Animals provided for in Regulation (EC) No 1760/2000,
  • The organic farming registers set up pursuant to Council Regulation (EC) No 834/2007,
  • Administrative sources associated with the cultivation of genetically modified crops and the specific rural development measures referred to in Annex III of the Regulation (EC) No 1166/2008.

Moreover, in case countries decide to use any other administrative source than those specified above, they can do so on condition that the Commission will be informed in advance and will be provided with a description of the used methods and quality of the data from this administrative source.

The article Farm structure survey - administrative sources provides a description of the administrative sources used in the FSS 2013 together with the main observations concerning the purpose and quality of these data sources, as well as the difficulties and measures taken to overcome them.

18.2. Frequency of data collection

The data collection is conducted every 3 or 4 years. See item 3.8.

18.3. Data collection

The methods used to collect the data vary in countries depending on national practices.

Detailed information for each country on the collection method used can be found in the national methodological reports.

18.4. Data validation

Data at Eurostat are checked by successive steps on:

  • the structure of the data set;
  • the internal relationship between fields (validation rules);
  • the raw aggregated results (control tables);
  • the cross-check with other agricultural statistics;
  • the cross-check with FSS data from previous years;
  • the applicability of precision requirements and whether precision requirements are met for the applicable cases;
  • the data analysis (outliers, distributions, weights etc.);
  • the review and validation of the national methodological reports.

For the complete overview of the validation workflow, see annex Validation workflow.

Only data and metadata having been entirely validated can be disseminated.
Afterwards some errors may be detected during a specific analysis and the data set revised.

Validation workflow
18.5. Data compilation

The aggregation can be simple summing up, but in case of sample surveys, the extrapolation factors provided by the countries for each holding are taken into account.

18.6. Adjustment

A few countries apply calibration (based on administrative sources) to improve the coverage and the data quality. For FSS 2013, see article Farm structure survey - administrative sources.

19. Comment Top

Not applicable.

Related metadata Top

Annexes Top