Farm structure (ef)

National Reference Metadata in ESS Standard for Quality Reports Structure (ESQRS)

Compiling agency: INE 

Time Dimension: 2016-A0

Data Provider: ES1

Data Flow: FSS_ESQRS_A

Eurostat metadata
Reference metadata
1. Contact
2. Statistical presentation
3. Statistical processing
4. Quality management
5. Relevance
6. Accuracy and reliability
7. Timeliness and punctuality
8. Coherence and comparability
9. Accessibility and clarity
10. Cost and Burden
11. Confidentiality
12. Comment
Related Metadata
Annexes (including footnotes)

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


1. Contact Top
1.1. Contact organisation
1.2. Contact organisation unit
Sub-Directorate General for Environmental, Agricultural and Financial Statistics
1.5. Contact mail address
Pº Castellana 183 

2. Statistical presentation Top
2.1. Data description
1. Brief history of the national survey 
The first Agrarian Census was performed in 1962, and was continued in the 1972 and 1982 censuses.

When Spain became a full member of the European Community on 1 January 1986, the INE joined the community programme of Surveys on the Structure of agricultural holdings. This programme requires changing the dates of the Agrarian Census, as determined by the subsequent Regulations of the Council, it must be conducted during the years ending in nine or in zero. The censuses have thus been carried out in 1989, 1999 and 2009, and the structural surveys in 1987, 1993, 1995, 1997, 2003, 2005, 2007 and 2013.


The survey for 2016 has two main objectives:

a) To evaluate the situation of Spanish agriculture and monitor the structural evolution of agricultural holdings, as well as to obtain comparable results from all the European Union Member States.

b) To comply with legal regulations set out by the European Union in the different Council regulations, as well as to meet national statistical requirements and other international requests for statistical information on the agrarian sector.


In order to meet these objectives and enable the comparativeness of the series, in general, the new 2016 maintains the scheme from the latest censuses and surveys.


2. Legal framework of the national survey 
- the national legal framework Unlike the case of the censuses, there is no specific legislation for these surveys.

This is due to the fact that, in accordance with Law 12/1989, of 9 May, on Public Statistical Services (LFEP), which regulates the statistical activity in Spain, any statistics listed in the National Statistical Plan are considered to be statistics for State purposes, and are mandatory.

Moreover, the second additional provision of Law 13/1996, of 30 December 1996, indicates that mandatory (compulsory) statistics are those whose performance is compulsory for the Spanish State, by requirement of European Union regulations.

The 2016 Survey on the Structure of Agricultural Holdings appears with programme number 6001, amongst the operations included in National Statistical Plan (NSP) 2013-2016, passed by Royal Decree 1658/2012, of 7 December, as well as in Royal Decree 1089/2015, of 4 December, passing the 2016 Annual Programme.

Furthermore, the FSS 2016 appears with programme number 7002, amongst the operations included in National Statistical Plan (NSP) 2017-2020, passed by Royal Decree 410/2016, of 31 October, as well as in Royal Decree 747/2016 of 30 December, passing the 2017 Annual Programme, of the aforementioned National Plan.

- the obligations of the respondents with respect to the survey The survey is compulsory, as in NSP 2013-2016 and NSP 2017-2020, it is considered statistics for State purposes.

Moreover, as the LFEP stipulates for all surveys and censuses, the main regulatory elements are included in the survey questionnaire: nature, characteristics and purpose of the survey, statistical secrecy, compulsory nature of supplying the data and sanctions set out for failure to comply with the law.

- the identification, protection and obligations of survey enumerators In accordance with the LFEP, and as statistics included in the NSP, the data is protected by Statistical Secrecy at all stages of compilation.

External company in charge of data collection and editing phases and all the personnel involved in the work are obliged to comply with the law:

- 12/1989 of 9 May, Law on the Public Statistical Function that requires the fulfillment of the duty of Statistical Secrecy and the information cannot be used by individuals or public or private entities, with the exception of INE.

- Organic Law 15/1999 of 13 December on Protection of Personal Data and the Regulation for its development, the Royal Decree 1720/2007.

- The personnel will be affected by the security measures established by Royal Decree, of 8 January, which regulates the National Security Framework within the e-government scope.

- The staff will agree to safeguard Statistical Secrecy in writing

- INE provides formal credentials for personal interviews.

2.2. Classification system

[Not requested]

2.3. Coverage - sector

[Not requested]

2.4. Statistical concepts and definitions
List of abbreviations
LFEP: (Ley de la Función Estadística Pública): Spanish law for Public Statistical Services.

NSP: National Statistical Plan.

EC:  European Community.

UAA: Utilised Agricultural Area.

LSU: Livestock Units.

SO: Standard Output.

TSO: Total Standard Output.

NMR: National Methodological Report.

IRIA: Integration of Information Collection and its Administration.

CAWI: Computer Assisted Web Interviewing

CATI: Computer Assisted Telephone Interviewing

CAPI: Computer Assisted Personal Interviewing

TAX ID NUMBER: Tax identification Number

PADRON: Continuous Municipal Register

EU: European Union

IACS: Integrated Agricultural Control System.

NUTS:  National Territorial Units for Statistics

FT: Farm Type

FT2: Farm Type (two digits).

ALL: Arable land+kitchen gardens+permanent crops.

GUAA: Group of UAA.

GALL: Group of ALL.

GLSU: Group of LSU.

AWU: Annual Working Unit.

FEADER:  European Agricultural Fund for Rural Development.

EUSTAT: Statistical Office of the Region of Pais Vasco (Autonomous Community).

L900: Free telephone number.

NE: Non-existent.

NS: Non-significant.

RSE: Relative Standard Errors.

ADI: Automatic Data Imputation.

IT: Information Technology.

2.5. Statistical unit
The national definition of the agricultural holding
The national definition of the holding is according to the EU definition fixed in the text of the Regulation (EC) Nº 1166/2008 and its annexes. No changes have been made.

The national definition of an agricultural holding is a unit, from the technical and economic perspective, with a single management, and which, within the Spanish economic territory, carries out agricultural activities, both as a main activity and a secondary activity. In addition, the holding may have other complementary (not agricultural) activities.
Said unit, on being a single unit from the technical and economic perspective, is characterised by a common use of labour and of means of production (machinery, land, installations, fertilisers, etc.). This implies that, if the plots of the holding are in two or more municipalities, they may not be very far from each other geographically.
The listing of agricultural and livestock activities is based on division 01 of the European Union Classification of Economic Activities (NACE, Rev.2), with some exceptions specified in Annex. In particular, it includes those holdings that maintain their land, which are no longer used for production purposes, in good agricultural and environmental conditions, in accordance with Council Regulation (EC) Number 1782/2003.
With the 2003 reform of the Common Agricultural Policy (CAP), the maintenance of the land in good agrarian and environmental conditions was introduced as an agricultural activity (Article 2 of said Regulation). Aside from this activity, farmers must not have any other agricultural activity in order to access the single-payment scheme.
Therefore, the agricultural holding may be defined as a unit with an agrarian nature (set of land and/or livestock), under a single management, located in a given geographical location, and which uses the same production methods. 

2.5.List of agricultural activities referred to in the definition of agricultural holding
2.6. Statistical population
1. The number of holdings forming the entire universe of agricultural holdings in the country

The framework population was 986,896 holdings (965,006 resulting from the last survey plus 21,890 holdings from the birth frame). To actualize the 2013 frame, we added the new holdings, which appeared after 2013 FSS, from administrative registers.


2. The national survey coverage: the thresholds applied in the national survey and the geographical coverage
The population surveyed in this survey is limited to holdings set down under article 3, namely, in Annex II of Regulation (EC) No. 1166/2008, which in our case concerns:

a) Agricultural holdings with at least 1 ha of utilised agricultural area (UAA). (A_3_1$ha)

A_3_1$ha>= 1

b) Agricultural holdings with at least 0.2 ha of UAA used for Fresh vegetables, melons and strawberries outdoor (B_1_7_1$ha) and flowers outdoors (B_1_8_1$ha); irrigated fruit and berry plantations (B_4_1$ha); irrigated citrus plantations (B_4_2$ha); nurseries (B_4_5$ha); permanent crops under glass (B_4_7$ha).

B_1_7_1$ha+B_1_8_1$ha+B_4_1$ha+B_4_2$ha+B_4_5$ha+B_4_7$ha >= 0.2

c) Agricultural holdings with at least 0.1 ha of UAA used for under glass fresh vegetables, melons and strawberries (B_1_7_2$ha)

B_1_7_2$ha >= 0.1

d) Agricultural holdings with at least 0.1 ha of UAA used for under glass flowers and ornamental plants (B_1_8_2$ha)

B_1_8_2$ha >= 0.1

e) Agricultural holdings with at least 0.5 ha of UAA used for tobacco (B_1_6_1$ha).

B_1_6_1$ha>= 0.5

f) Agricultural holdings with at least 0.5 ha of UAA used for hops (B_1_6_2$ha).

B_1_6_2$ha >= 0.5

g) Agricultural holdings with at least 0.5 ha of UAA used for cotton (B_1_6_3$ha).

B_1_6_3$ha >= 0.5

h) Agricultural holdings with one or more livestock units (LSU) and a total standard output (TSO) equal to or above 900 Euros.

These criteria are independent: at least one must be met for a holding to be eligible for the Survey.

The survey covered the whole territory of Spain.

These thresholds were already used in last Census and 2013 FSS.


3. The number of holdings in the national survey coverage 
The framework population was 986,896 holdings (965,006 resulting from the last survey plus 21,890 holdings from the birth frame). To actualize the 2013 frame, we added the new holdings, which appeared after 2013 FSS, from administrative registers.

The initial sample was 66,051 units and the number of holdings which answered was 56,051.

The number of holdings in the final weighted population of the FSS 2016, according to the national survey coverage, is 945,024 holdings.


4. The survey coverage of the records sent to Eurostat
The survey coverage of the records sent to Eurostat is the same as the national survey coverage. 


5. The number of holdings in the population covered by the records transferred to Eurostat
The number of holdings in the population covered by the records transferred to Eurostat is 945,024.


6. Holdings with standard output equal to zero included in the records sent to Eurostat
In the file sent to Eurostat there are 604 records with standard output equal to zero which represent 11 286 in the elevated population.

All these 604 holdings are eligible. They have other livestock, permanent grassland no longer in production purposes, and the most have fallow land.

The permanent grassland no longer in production purposes is kept in good agricultural and environmental conditions, so the holdings are eligible for the survey.

Fallow land is a practice of cultivation used to improve the quality of the soil and the better use of rain water. They are part of the crop rotation and are taken out of production usually for a year but they will be cultivated in the next agricultural period.


7. Proofs that the requirements stipulated in art. 3.2 the Regulation 1166/2008 are met in the data transmitted to Eurostat
We don't change the threshold used in Census 2009. Therefore, it is valid the argumentation described in the Census NMR:

Prior to fix the conditions of the threshold, we assured that it excluded the smallest holdings which together contribute 2% or less to the total UAA (excluding common lands) and 2% or less to the total number of LSU. According to the last Agricultural Census, only 0.66% of the UAA (excluding common lands) and 0.44% of the total number of LSU are excluded.

See attached file Statistics on holdings above and under the thresholds.


8. Proofs that the requirements stipulated in art. 3.3 the Regulation 1166/2008 are met in the data transmitted to Eurostat
The survey uses lower thresholds than in Annex II of Regulation 1166/2008.

2.6-7. Statistics on holdings above and under the thresholds
2.7. Reference area
Location of the holding. The criteria used to determine the NUTS3 region of the holding
To determine the location of each holding in one of these areas, the province and municipality to which the holding is allocated were used, which is the one in which most of the holding is located and if it doesn't have land, where its livestock is declared. 
2.8. Coverage - Time
Reference periods/dates of all main groups of characteristics (both included in the EU Regulation 1166/2008 and surveyed only for national purposes)
The reference periods of the data are in accordance with community regulations.
  • For the characteristics relating to the land and labour, the reference period is agricultural year 2016, that is, the agricultural campaign between 1 October 2015 and 30 September 2016.
  • For the head of livestock, the reference date is 30 September 2016.
  • For rural development measures, the reference period is the last two years, that is, from 1 January 2015 to 31 December 2016.
2.9. Base period

[Not requested]

3. Statistical processing Top
1.Survey process and timetable
Calendar (overview of work progress)

- Constitution of the working group for preparing and carrying out the project:

Starting time: March-2015.

Ending time: December-2017

This working group comprises all the units involved in the project: Responsible Department, Secretariat General, Sampling Unit, Data Collection Unit, Statistical Information Technology and Statistical Dissemination.

- Inclusion in the annual budgets for 2016 and subsequent years: March 2015

- Analysis of the requirements of the Regulation, in terms of: questions from the questionnaire, sampling errors, collection periods, variation in farm type to be researched: March 2015

- Initial proposal of the collection method. Decision with regard to the use of IRIA software (Integration of Information Collection and its Administration). Alternatives with regard to externalising or internalising some collection phases: March 2015.

- Calculation of an approximate sample size. Decision with regard to increasing the sample, based on the possible incidences, rather than processing via replacements: March 2015

- First budget: April 2015.

- Preparation of a detailed calendar of tasks: March 2015

- Presentation of the Grant application to Eurostat: 20 March 2015. (an amendment was presented on 10 August 2015)

- First draft of the questionnaire: June 2015

- Contacts with the Statistics Institutes of the Autonomous Communities: October 2015

- Writing of the partnership agreement with the Statistics Institute of País Vasco: March 2016 - 15 June 2016

- Preparation of the validation rules (edits) for the collection phase: November-December 2015.

- Sample selection: December 2015.

- Preparation of the methodological project: June-December 2015

- Questionnaire and Instruction sheet, print version: September-November 2015

- Specifications of the IRIA application for collection via each channel: CAWI, postal, CATI, CAPI: October 2015 to March 2016.  

- Design of the results tables: October-December 2015

- Specifications for the application of the centralized filtering following the data collection: January-March 2016

- Preparation of the guidelines for the hiring of an external company for the data collection: August 2015-November 2015

- Writing of the letters for the initial contact with the respondents, and for claims: March 2016.

- Translations of the questionnaire, instruction sheet and letters into the co-official languages: March-April 2016

- Development and tests of the IRIA application: March-September 2016

- Preparation of the training manual for the survey content: April 2016

- Preparation of the training manual for IRIA: August-September 2016

- Uploading of the sample in IRIA for phase 1 of the collection (Postal, CAWI and CATI): September 2016

- Assessment and resolution of the award contract of the data collection:

Assessment:  5 May – 26 May 2016

Award: 9 August 2016

Formalization of the contract: 4 October 2016

- Personnel training of the awarded company: 25 and 27 October 2016

- Information collection in phase 1 (Postal, CAWI and CATI) and inspection of the collection and monitoring: 28 November 2016 - 1 March 2017

- Information collection by telephone interviews (CATI): January 2017 - 1 March 2017

- Information collection in phase 3 (CAPI) by the external company, and supervision and monitoring by Central Services: March – 20 May 2017

 - Provisional calculation of the elevation factors: June 2017

- Analysis of non-response and identification of relevant units in the estimations for performing the additional attempts for the collection or confirmation of the information: January-June 2017

- Development and tests of the application for centralised filtering, subsequent to the data collection: March-September 2016

- Filtering of the collection by the Responsible Department: December 2016-December 2017

- Register design for the transfer of the information from the IRIA application, to the subsequent centralised processing: January-March 2016.

- Specifications of the rules for automatic imputation: January 2016-October 2017.

- Programming, tests and performance of the automatic imputation. April 2016- November 2017. 

- Calculation of the final elevation factors: October- November 2017.

- Specifications for other subsequent processing and results tabulation rules: January 2016- November 2017.

- Programming, tests and execution of other subsequent processing and result tabulation rules: June 2016-November 2017.

- Specification of the norms for obtaining the EUROFARM file: October 2016-October 2017.

- Programming and execution of the norms for obtaining the EUROFARM file: November 2016- November 2017.

- Preparation of the methodological report of the Survey: September 2017-November 2017

- Mailing of microdata to Eurostat: December 2017

- Publishing of the results of the Survey: December 2017


Personnel that have worked on the project in the Units that have collaborated with the Survey

The annex "Involved personnel" shows the number of persons who worked in each of the Units in the statistical operation, broken down by different categories. The subdirectors of each Unit have not been included. This personnel has not necessarily worked full-time on the Survey. 


2. The bodies involved and the share of responsibilities among bodies
Responsible Department

- Preparation of the methodological project

- Draft questionnaire

- Establishment of the validation norms to be passed during the collection phase

- Design of the results tables

- Specifications for the application of centralised filtering subsequent to the data collection

- Filtering of the collection data by the Responsible Department

- Specifications of the rules for automatic imputation

- Specifications for other subsequent processing like the way to calculate aggregate variables (Annual Working Units, Livestock Units, Total Standard Output and Farm Type), other instructions and results tabulation rules

- Specification of the norms for obtaining the EUROFARM file


Data collection

- Preparation of the guidelines for the hiring of an external company for the collection

- Writing of the letters for the initial contact with the respondents, and for claims

- Translations of the questionnaire, instruction sheet and letters into the co-official languages                                                                                

- Preparation of the training manual for the survey content

- Monitoring and control of the tasks of the external company

- Inspection of the CAPI collection and supervising questionnaires completed in any channel 


Sampling Unit                                             

- Sample selection

- Calculation of the elevation factors and the sampling errors


Information Technology Unit

- IRIA training manual

- Uploading of the sample in IRIA

- Programming, testing and execution of the automatic imputation

- Programming, testing and execution of other subsequent processing like calculate aggregate variables (Annual Working Units, Livestock Units, Total Standard Output and Farm Type)

- Programming, testing and execution of  result tables

- Programming, testing and execution of the norms for obtaining the EUROFARM file

- Mailing of microdata to Eurostat


External company

-Design and order of envelopes for the postal collection

-Training courses for interviewers and interviewer inspectors

- Information collection: sending, recording and initial filtering of questionnaires; conducting of interviews, by telephone or personal; and inspection of responses completed in phase 1 (Postal, CAWI and CATI).

- Partial data editing, check and corrections between data collection and the final statistical analysis. The rest of the data editing was performed by Responsible Department staff.


Secretariat General and Data Collection:

- Annual budgets for 2016 and subsequent years 


Responsible Department and Cabinet

- Contact with the Statistics Institutes of the Autonomous Communities

- Writing of the partnership agreement with the Statistics Institute of País Vasco


Data collection and Information technology

- Specifications of the IRIA application for collection via each channel: Postal, CAWI, CATI and CAPI.

- Initial proposal of the collection method. Decision with regard to the use of IRIA.  


Responsible Department and Information Technology

- Development and tests in the application of centralised filtering subsequent to the data collection


 Responsible Department, Data Collection and Dissemination

- Publishing and design of the information leaflets


 Responsible Department, Data Collection and Sampling 

- Analysis of the requirements of the Regulation, in terms of: questions from the questionnaire, sampling errors, collection periods, variation in farm type to be researched

- Sample size and incidence processing 

- Questionnaire and Instruction sheet, print version


Responsible Department, Data Collection and Information Technology

- Assessment and resolution of the award contract of the collection

- Personnel training of the awarded company


Responsible Department, Information Technology and Dissemination

- Publishing of the results of the Survey


Responsible Department, Information Technology, Data Collection and Sampling

- Preparation of the Survey quality report


All units

- Establishment of the detailed calendar of tasks

- Preparation of documentation for the Grant application for Eurostat


3. Serious deviations from the established timetable (if any)
There was a deviation of 3 months in the starting of the data collection due to financial matters. In spite of this delay, the final data were in time.

3-1.Involved Personnel
3.1. Source data
1. Source of data
The Farm Structure Survey 2016 has been carried out as a survey in which a questionnaire was sent out to each unit of the sample. All variables apart from personality of the holding, age and sex of the holders were included in the questionnaire.

The variables of sex and age of individual farm holders were obtained by using the Tax ID Number to cross-reference the data collected in the survey with the PADRÓN (Continuous Municipal Register). We obtained the personality of the holding through the ID number (that it is not a register) by a set of established rules.


2. (Sampling) frame
Initially, the sampling framework for 2013 FSS was the listing of agricultural holdings obtained in the 2009 Agricultural Census, which meets the requirements of the survey with 989,796 holdings.

This sampling framework is available after conducting a census (every 10 years).  It is updated with the FSS surveys, through the procedures of the subsidiary operations.

In 2016 FSS we return to investigate 2013 sample (it’s a panel). Moreover, we form a birth frame with the new holdings (later on 2013) from administrative files (mainly on agriculture characteristics, known as IACS and on livestock, known as Livestock Register). Both of them are updated annually.


3. Sampling design
3.1 The sampling design
One-stage stratified random sampling of agricultural holdings
3.2 The stratification variables
For the panel sample, strata were formed by NUTS2 x two-digit farm types (FT2) x 6 size groups.

To build the size groups, we consider UAA and ALL (arable land + kitchen gardens + permanent crops) as variables for stratification in predominantly agricultural FT2s and UAA and LSU in predominantly livestock FT2s.

For each stratification variable, we apply the Dalenius and Hoges (1959) or the cumulative square root rule (Cochran, Sampling Techniques, Mexico, 1980, pp. 169-174). We set 5 strata and build categorical variables, GUAA, GALL and GLSU, which take  values between 1 to 5; 1 corresponds to smaller holdings and 5 corresponds a larger holding. Then, the size group definition is given as following: 

For predominantly agricultural FTs:

  •  Size=MAX(GUAA,GALL)

For predominantly livestock FTs:


To reach more homogeneity with respect to stratification variables, in the stratum with the largest holdings, we define an additional size. If both categorical variables take the value 5, then the size group takes the value 6.

At the end, in each NUTS2xFT2, we have 6 size groups. The size group equal to 7 corresponds to holdings from take-all stratum (see 3.1-3.3 below).

For the new sample, obtained from birth frame, we also build strata according to NUTS2 x crops/livestock type x 4 size groups. According to type, we have 6 types for livestock and 8 types for crops. The way to build the size group is similar to the other case: we apply cumulative square root rule to two variable: total crops and total LSU.

3.3 The full coverage strata
For the panel sample, firstly, the exhaustive holdings are those that meet one of the following conditions:

UAA >= 5000 Ha, or ALL>= 1000 Ha or LSU >= 5000 or AWU >=50. Depending on the characteristics of each region (NUTS2), these limits are reduced.

Secondly, we determine exhaustive holdings using the sigma deviation rule (Julien and Maranda, Le plan de sondage de l'enquête nationale sur les fermes de 1988. Techniques d'enquête 1990, vol.16 nº1). This rule is applied in each group formed by NUTS2xFT2, to the variables UAA, ALL, LSU and AWU.

For the new sample, we use the same technique in each group formed by NUTS2xtype, to the variables total crops and total LSU.

These exhaustive holdings go to the 7 size group.

3.4 The method for the determination of the overall sample size
The overall sample size is 65,652 holdings. 

The difference between 65,652 and 66,051 (Number of holdings in the gross sample plus possible (new) holdings added to the sample) is due to daughter holdings. The used methodology is explained in point "3.6 Sampling across time design".

Daughter holdings are collected during field work. For this reason, they are not taking into account in the initial sample size.

The overall sample size is obtained as a result of multivariate optimum allocation as it is specified in point 3.5

3.5 The method for the allocation of the overall sample size
For the panel sample, an optimum multivariante allocation, using the precision requirement established in Annex IV of Regulation EC No 1166/2008, was applied. It was resolved using the Bethel algorithm (Répartition de l'échantillon dans les enquêtes à plusieurs variables, Techniques d'enquêtes, 1989, vol.1 nº 1, pages 49-60).

For the new sample, we use an optimum and proportional allocation. 

3.6 Sampling across time
Farm Structure Surveys are conducted using a panel. In FSS 2013, a sample was obtained from the directory available at the time (Agricultural Census 2009) and the same sample has been investigated in 2016, updated as follows:

-For holdings with land, with the new holdings obtained from the survey using the mother and daughter farm methodology.

-For a sample obtained from birth frame with null intersection with the previous frame.

3.7 The software tool used in the sample selection
Tailor-made programs using SAS software. 
3.8 Other relevant information, if any
Not available.


4. Use of administrative data sources
4.1 Name, time reference and updating
PADRON (Municipal Continuous Register) is an administrative dataset managed by the INE that is continuously updated by town and city councils. Its purpose is to provide the official population figures, approved by Royal Decree, of all Spanish municipalities at 1 January each year. It contains a list of all residents with the following variables, among others: Tax ID No., sex, age, place of birth, place of residence (with full postal address) and nationality. Year 2016

IACS register (Council Regulation (EC) No 73/2009). To update the sample we used IACS register 2015 and to check the results we used IACS 2016. 

Register of livestock holdings (Royal Decree 479/2004 of 26 March,establishing and regulating the General Register of livestock farms, Ministry of Agriculture ). We use the livestock register 2015 to update the sample and we use livestock register 2016 to check the results. 

Rural Development Programmes. The 2015-2016 rural development programming is applied in Spain according to their competence framework and , therefore, in addition to the relevant National Strategic Plan required under Title III, Chapter 1 of Regulation ( EC) 1305/2013 on support for rural development through the "Fondo Europeo Agrícola para el Desarrollo Rural, FEADER" (European Agricultural Fund for Rural Development) , there are seventeen regional programs , one for each region. We used 2015 and 2016 versions to check the results. 

4.2 Organisational setting on the use of administrative sources
The Ministry of Agriculture has provided INE with data from IACS, Animal Register and others. 
4.3 The purpose of the use of administrative sources - link to the file
Please access the information in the file at the link: (link available as soon as possible)


4.4 Quality assessment of the administrative sources
  Method  Shortcoming detected Measure taken
- coherence of the reporting unit (holding)   The units in administrative sources do not correspond directly to the definition of required statistical units.  
- coherence of definitions of characteristics   The variables of the administrative registers cannot be used directly in FSS.  
- coverage:      
  under-coverage PADRON is an administrative dataset which contains information about all residents in Spain. We obtained the age and sex of individual farm holders for more than 98% of the holdings by means of the cross between PADRON and 2016 FSS.

PADRON: A small percentage of holdings was not covered by this source.

PADRON: For these holdings, we used the information from 2009 Agricultural Census data to obtain sex and age of the holder. 
  multiple listings      
- missing data      
- errors in data      
- processing errors      
- comparability      
- other (if any)      


4.5 Management of metadata
Not relevant.
4.6 Reporting units and matching procedures
PADRON: residents.
The Tax ID Number (a unique 9-position alphanumeric code) of the holders was used to obtain the legal personality of the holding, the sex and the age of holders.

IACS register: farmers who receive subsidies;

Register of livestock holdings: farms with cattle, sheeps and goats, equines, pigs, poultry, rabbits, hives and other chase animals;

Rural Development Programmes: holders which receive subsidies from European Agricultural Fund for Rural Development.

4.7 Difficulties using additional administrative sources not currently used
Absence of a key variable to do the crosses.
3.2. Frequency of data collection
Frequency of data collection
The frequency of data collection is established in the Regulation Nº 1166/2008 
3.3. Data collection
1. Data collection modes
The INE signed a partnership agreement with the Autonomous Community of País Vasco, by which, its Statistics Institute (EUSTAT) performs the fieldwork and the recording of all the questionnaires in its territory.

In the information collection performed by the INE, a multi-channel methodology was used, with different collection systems:  by postal, online completion (CAWI), computer-assisted telephone interviewing (CATI) and computer-assisted personal interviewing (CAPI). It was carried out in two differentiated phases, over the course of six months.

An external company hired through public tender by INE was in charge of performing all phases of data collection.

  • Phase 1: Postal, Computer-assisted web interviewing (CAWI)  and Computer-assisted telephone interviewing (CATI)  phase

This phase began by mailing the questionnaires, by ordinary post, to the owners of the agricultural holdings in the sample that would complete the print questionnaire and return it by post. Simultaneously, the CAWI channel was enabled, permitting completion of the questionnaire online, without the direct involvement of an interviewer, in such a way that the interview interacted directly with the system. In both cases, the collection system is the individual completion by the respondent.

After the beginning of the survey, two letters were sent to non-respondents. Once the second letter was sent, it began the collection by telephone interviews, channel CATI, of those owners of agricultural holdings that had not returned the completed questionnaire and with available telephone numbers.  

As support for the collection, and in order to answer calls from respondents, a free telephone lines was provided (L900).

The questionnaires received by Post or CAWI during this phase have been recorded, revised and filtered, contacting - by telephone - those respondents for whom some sort of clarification or correction was required regarding the data provided.

This phase was carried out from 28 November 2016 to 1 March 2017. It began with the mailing of letters and questionnaires to all of the owners of the agricultural holdings in the sample, together with the user keys and passwords, in order to be able complete it online. The telephone interviews began on January 2017 after two claim letters were sent. The CAWI channel was closed on 1 March 2017. The Postal and CATI channels were maintained up until the end of the collection.

  •  Phase 2: Computer-Assisted Personal Interviewing (CAPI)

This collection channel allows for personal interviewing, following the questionnaire script programmed into a portable device.

This procedure was reserved for those owners of agricultural holdings for whom information had not been obtained in the previous phase.

This phase was carried out from March to 5 May 2017. The INE inspection of CAPI phase was ended on 20 May 2017. 


2. Data entry modes
The questionnaires received by post were recorded and filtered by the personnel in charge of their collection.

In the case of the questionnaires that entered via the CAWI channel, it was the holders themselves who were in charge of introducing the data. These questionnaires were subject to a subsequent filtering in the collection unit.

Lastly, the interviewers directly recorded the interviews conducted in the CATI and CAPI channels. In the case of the personal interviews, portable devices were used to collect the data.

The computer application (IRIA) for performing the recording (by the holders or the interviewers) and filtering work was the same for all of the channels.

Filtering work consisted in the application of validation edits (coherence between characteristics of the questionnaire) and flow edits.


3. Measures taken to increase response rates
  • Phase 1: Postal, CAWI and CATI  phase

During this first collection phase, the mailing of the questionnaire by post or CAWI, as many as two letters were sent, staggered during the first two months of the collection period, requesting completion of the questionnaire. 

 Once the telephone interviews, CATI channel, began, several calls were made in order to contact non-respondents and achieve their collaboration. This gave priority to the collection through this means of holders with several holdings, and in populations with few respondents, in the hopes of avoiding dissemination in the CAPI phase as much as possible.

  • Phase 2: CAPI phase
In conducting the personal interviews, the CAPI channel, prior to the interview, a visitation letter was sent, with the date and place of the interview (home address of the holder). These appointments were confirmed by telephone. In case of holders or respondents' absence on the established date, a notice for a new visit was provided. If it was not possible to contact the holder after several attempts, a new visit letter was sent. It was established to make at least six visits to the address, in order to conduct the interview.


4. Monitoring of response and non-response
1 Number of holdings in the survey frame plus possible (new) holdings added afterwards

In case of a census 1=3+4+5

2 Number of holdings in the gross sample plus possible (new) holdings added to the sample

Only for sample survey, in which case 2=3+4+5

3 Number of ineligible holdings 3325 
3.1 Number of ineligible holdings with ceased activities

This item is a subset of 3.

4 Number of holdings with unknown eligibility status


4.1 Number of holdings with unknown eligibility status – re-weighted 0
4.2 Number of holdings with unknown eligibility status – imputed 0
5 Number of eligible holdings


5.1 Number of eligible non-responding holdings


The difference of 538 eligible non-responding units which are not imputed nor re-weighted is that they belong to exhaustive strata, without information for an imputation, and only for those cases we have considered more convenient neither to impute nor re-weight.

5.1.1 Number of eligible non-responding holdings – re-weighted 2898 
5.1.2 Number of eligible non-responding holdings – imputed 3239 
5.2 Number of eligible responding holdings 56051 
6 Number of the records in the dataset 




5. Questionnaire(s) - in annex
See annexes.

3.3-5. Questionnaire FSS 2016
3.3-5. Additional questionnaire (labour force and other gainful activities directly related to the holding)
3.4. Data validation
Data validation
In the IRIA application, approximately 135 groups of edits have been programmed, which are distributed among the following categories:
  • Data format cheks: 10
  • Completeness checks: 45
  • Routing checks: 35
  • Outliers checks: 5
  • Relational checks: 35
  • Ratio checks: 5

IRIA Software (Integration of Information Collection and its Administration) was the tool used during data validation.

Three information validation levels must be distinguished:

Validation during the information collection: there were controls included in the IRIA application in each one of the collection phases (CAWI, post, CATI or CAPI). The controls were presented to the interviewers either during the interview itself, or at the end thereof, according to the type of error. In the case of the postal collection, in which the questionnaire was recorded once it had arrived by post, the controls were detailed at the end of the interview, and resolved via telephone calls. Subsequently, the interviewer inspectors were required to accept or reject each of the questionnaires from the interviewers, depending on the types of error contained, and on the observations included on them. At the following level, the survey inspector performed a global inspection of the information collected.

Validation in the Central Services: once the questionnaires were marked as cleaned in the collection phase, the Responsible Department performed an information validation, mainly guided for the identification of the observations to be dealt with, based on the coherence in the evolution of the estimated data, with regard to the results available from previous Surveys or from the census. Likewise, a monitoring was performed of the incidences in the collection, and specific research regarding some of them, as well as regarding the daughter holdings.

Automatic imputation: an automatic imputation process was performed on the information, using a specific programme.

3.5. Data compilation
Methodology for determination of weights (extrapolation factors)
1. Design weights
The sampling weights are determined as the inverse of the probabilities of selection. 
2. Adjustment of weights for non-response
The weights are adjusted by non-response in the cases indicated in the previous point. The resulting adjusted estimator is the re-weighted Horvitz-Thompson estimator. Homogeneous response groups are not created. The adjustment due to non-responses is performed at the stratum level.
3. Adjustment of weights to external data sources
No adjustments of weights to external data sources were made. 
4. Any other applied adjustment of weights
In editing process, some stratum changes were performed and the sampling weights were adjusted. 
3.6. Adjustment

[Not requested]

4. Quality management Top
4.1. Quality assurance

[Not requested]

4.2. Quality management - assessment

[Not requested]

5. Relevance Top
5.1. Relevance - User Needs
Main groups of characteristics surveyed only for national purposes 
The characteristics researched are reduced to the list proposed in Community Regulation EC (No.) 715/2014.

Due to the importance of irrigation in Spain, because of its effect on the crop yields and on the fact that water is a scarce commodity, with an uneven geographical distribution, the non-irrigation and irrigation lands of all crops are researched separately.

It is essential to ascertain the irrigated areas of each crop, in order to understand Spanish agriculture, and this is required by most national users (Ministry of Agriculture, producer associations, researchers, individuals, etc.).

5.2. Relevance - User Satisfaction

[Not requested]

5.3. Completeness
Non-existent (NE) and non-significant (NS) characteristics - link to the file. Characteristics possibly not collected for other reasons
Please access the information in the file at the link: (link available as soon as possible)
5.3.1. Data completeness - rate

[Not requested]

6. Accuracy and reliability Top
6.1. Accuracy - overall
Main sources of error
We calculate sampling errors of the main variables and we analyze the fulfilment of precision's requirements established in Regulation 1166/2008. All requirements are fulfilled except for the variable "Breedings sows" in two NUTS2 and one NUTS1 and the variable "Area of citrus plantation" in one NUTS2.

To analyze the non sampling errors, we studied the non-reply in every one of the strata obtained by the cross between NUTS2, farm type and size strata. Imputation and re-weighting processes are applied for their treatment. External sources are not used. The high level percentage of reply, more than 89,4%, and the procedures applied in the treatment of non-reply lead to a reduction of the possible bias.

6.2. Sampling error
Method used for estimation of relative standard errors (RSEs)
We use simple expansion estimators and for variance estimation, we apply the Raulin formula. Formulas are provided in the attached file 6.2. Method used for estimation of relative standard errors (RSEs)

6.2. Method used for estimation of relative standar errors (RSEs)
6.2.1. Sampling error - indicators

1. Relative standard errors (RSEs) - in annex


2. Reasons for possible cases where precision requirements are applicable and estimated RSEs are above the thresholds
The relative standard error of the amount of "breeding sows" errors for ES24 is equal to 7,45% and for ES51 is 5,5%.

The sample was obtained having into account the variable "breeding sows" in the precision requirements for each considered NUTS. The variability of that variable and the changes in time have originated that the final error exceed 5% in the resultant sample.

For NUTS1= ES2, the relative standard error of the amount of breeding sows is equal to 5,96%. This result is due to the ES24, which is a part of ES2.

In the case of “Area of citrus plantation” in NUTS2= ES62, the relative standard error is equal to 5,48%.

As the preceding case, the sample was obtained taking into account such variable in the precision requirements. Changes in time have originated that in the resultant sample the RSE exceed the 5%.

6.2.1. Relative standard errors
6.3. Non-sampling error

See below

6.3.1. Coverage error
1. Under-coverage errors
Not available


2. Over-coverage errors
Over-coverage, measured through non-eligible units, accounts for 5.1% of the initial sample. The estimated data are corrected, due to over-coverage, calculating a deflation factor in the calculation of the size of the estimated population.
2.1 Multiple listings 
Duplicated holdings represent 0.4% of the total sample. These holdings reduce the population, as do the rest of the holdings that cause over-coverage.


3. Misclassification errors
We try not to change the initial strata to respect the initial selection probabilities. Stratum changes are only performed in those influential units.

The procedure generally following in the stratum changes is that the holding that changes goes to the new stratum, maintaining the elevation factor of the stratum of origin. When the jump is large, for example from stratum size 1 (small holdings) to stratum size 6 (large holdings), the factor of the changing holding takes the value 1.


4. Contact errors
Non-contact holdings are holdings which have been impossible to contact them after sending questionnaire by mail and trying to call them. We treat to the non-contact holdings like eligible holdings because they come from our agriculture frame. This frame has been built using different agriculture sources. If we don't receive any information about their cease, we apply re-weighted or imputation procedures. 


5. Other relevant information, if any
 Not available. Over-coverage - rate
Over-coverage - rate
Over-coverage rate 5.0 % Common units - proportion

[Not requested]

6.3.2. Measurement error
Characteristics that caused high measurement errors
The principal causes of measurement errors are due to the self-completion without the help of interviewers. 

We improved the questionnaire and the collection method with the experience acquired with the Census and 2013 FSS. For instance, we included in all questionnaires the possibility of using different units of measurement from hectares and we eliminated the possibility of having more than one questionnaire per holding (paper, CATI, CAWI, CAPI).

To ensure the coherence of data and to minimize the errors, we used an application (IRIA) performed by INE which integrated all the phases of collection and editing of data.

All questionnaires (paper, CAWI, CATI and CAPI) were recorded using IRIA.

During the collection and recording phases for mailed questionnaires, the data underwent a check, with a quality control of recording and control of the data supplied.

Also, CAWI, CATI and CAPI have their own controls in IRIA.

IRIA detects mistakes in the internal consistency of questionnaires (partial absence of data in a questionnaire, inconsistent data between different variables and control of the range and the existence of quantitative variables). It also detects and lists controls of outliers, such as crops which are no common in certain regions. In all, 96 controls were set up.

A post-recording editing was performed centrally by the Promoting Unit, helped by an external enterprise which provided a team of editors for the task using IRIA as well. After this manual correction of errors  and prior to obtaining the datasets with the final data, all questionnaires underwent a process of Automatic Data Imputation (ADI).

6.3.3. Non response error
1. Unit non-response: reasons, analysis and treatment
The main reasons for non-response are:

- It has been impossible to obtain the questionnaire, it means there had been some contact with the holding, we knew there had been a change in the holding but at the end, it wasn't possible to collect the information.

- Refusals: it means that the only contact was a 'NO'.

- Non-contact: these holdings have been impossible to contact.

We decided to impute non-response caused by refusals or non-contact. In the cases where there was a first contact but at the end, it wasn't possible to obtain the questionnaire, we applied re-weighting.


2. Item non-response: characteristics, reasons and treatment
Some characteristics like Labour Force have a high item non-response rate due to the difficulty and the excessive detail of this block of variables.

Regarding to their treatment, you can see 6.3.4. Processing error - items 1. Imputation methods and 3. Tools used to make corrections. Unit non-response - rate
Unit non-response - rate
According to the status collection, the non-response rate is 10.6%. Item non-response - rate
Item non-response - rate
The characteristics related to family labour force (sex, other gainful activities, etc.) have a high item non-response but individual data for this item is not available. 
6.3.4. Processing error
1. Imputation methods
Following the manual correction of errors and prior to obtaining the datasets with the final data, all questionnaires underwent a process of automatic data imputation (ADI).
If there are no inconsistencies, the block makes no imputations to the holding in question and moves on to the next holding. If a block applies imputations to a holding, the amended data are final; therefore, queries made by subsequent blocks refer to the updated rather than the original data. This also applies within each block. Queries made after application of one or more imputations always refer to the updated status of data, even where modified within the process of that same block.
Imputations are of two general types. One type is deduced from the information in the questionnaire itself by applying given criteria, while the second type requires recourse to external information to make up for missing data in the questionnaire. Imputations of the first type, where they relate to arithmetic inconsistencies, squaring up sums for instance, generally operate by imputing new data in proportion to those appearing in the questionnaire the sum of which verifies the desired consistency condition. Imputations of the second kind refer to information drawn from external sources usually the last available data (Census 2009 and FSS 2013). 


2. Other sources of processing errors
The number of holdings affected with errors was 19 828.

In post-recording editing, when the editor detected the error, he or she could change the questionnaire data through IRIA. This application also allowed editors to bring up a scanned image of the questionnaire, where available, to assist with the editing process.

Editing was performed as follows:

- Firstly, all the holdings outside the set threshold were investigated again to confirm that they were indeed outside the threshold.
- Secondly, holder IDs were corrected (no Tax ID No. or incorrect Tax ID No.). For this step, the editors did a manual search in PADRÓN for 100 Tax ID Numbers using names and surnames.
- Lastly, the holdings resulting from the above process underwent the editing process. A check on 96 errors was carried out in each questionnaire to detect data inconsistencies or outliers. 


3. Tools used and people/organisations authorised to make corrections
IRIA was programmed by the IT Unit of INE. This application will be used in other surveys of INE.

The automatic data imputation (ADI) phase was programmed entirely by the IT Unit of INE based on the specifications of the Promoting Unit.
The ADI comprises fourteen blocks, each of which performs a specific function. Blocks are applied in the sequence 1 to 14 to each holding; each block basically conducts three types of operation:
- Queries to detect inconsistencies.
- Queries to acquire information from the questionnaire itself where inconsistencies have been detected.
- Imputations as necessary. Imputation - rate
Imputation - rate
The imputation phase affected 8.7% of the gross sample (66 051 holdings) and it includes all inconsistencies between different items of the questionnaires, items “without data” (item non-response) or items with “no valid data”. 
6.3.5. Model assumption error

[Not requested]

6.4. Seasonal adjustment

[Not requested]

6.5. Data revision - policy
Data revision - policy
Only the final data of the Survey is published, and it is not subject to revision.

If errors are detected and the data needs to be modified, an explanatory note would be added to the information in order to inform users that the data has been changed.

6.6. Data revision - practice
Data revision - practice
The variables of the table in section 8.2. Comparability - over time - item 6. were subject to special analysis at various levels of regional disaggregation: National level and Autonomous Community. This has helped to achieve the same quality as previous Surveys.

Data have been revised throughout all the process.

Weekly, all data were revised to check the changes made in the week and to compare the provisional obtained results with the previous data (last FSS) and Ministry Agricultural data. For that, data from IRIA were downloaded and tabulated to make comparisons. 

Only the final data of the Survey is published, and it is not subject to revision.

If errors are detected and the data needs to be modified, an explanatory note would be added to the information in order to inform users that the data has been changed.

6.6.1. Data revision - average size

[Not requested]

7. Timeliness and punctuality Top
7.1. Timeliness

See below

7.1.1. Time lag - first result
Time lag - first result
No interim results have been published. 
7.1.2. Time lag - final result
Time lag - final result
The time lag for the final survey results is 14 months and 13 days. The reference moment used for the calculation of the time lag is 30 September 2016 because is the last day of the agricultural campaign in Spain.
7.2. Punctuality

See below

7.2.1. Punctuality - delivery and publication
Punctuality - delivery and publication
They were no delays in the publication of the results. 

8. Coherence and comparability Top
8.1. Comparability - geographical
1. National vs. EU definition of the agricultural holding
There are not differences between the national definition and the EU definition of the holding. 


2.National survey coverage vs. coverage of the records sent to Eurostat
The population covered in the national survey is the same as the population covered by the records sent to Eurostat. 


3. National vs. EU characteristics
As the design of the questionnaire (prior to the print version of it) was finished in November 2015, the definitions of characteristics have been based on versions of the Handbook on implementing the FSS prior to that date. We can conclude therefore that the methodology used does not differ to that of the EU.

The number of hours per year for a full-time employee is 1824.


4. Common land
4.1 Current methodology for collecting information on the common land
Common Lands in Spain are usually permanent grassland (2.03.01+2.03.02) used as pasture for cattle+ lands not forming part of the UAA [wooded area (2.05.02) + other lands (2.05.03)]. In most cases, Arable land and permanent crops are not part of Common Lands.
Common land area is only counted once.

In the case of common land used jointly by several holdings, since it is not possible to assign a specific section to each farmer, the common land is considered a separate holding and all the land (without the cattle grazing on it) is counted in that holding, as with any other. The relevant common or local authority (State, Autonomous community, neighbourhood community, parish, etc.) is listed as owner of the holding.
If during the agricultural production year, the owner leases or freely assigns all or part of the land to a single holding, the transferred/leased part is allocated to the holding that individually works this land.

The type of tenure of the common land assigned to holdings is "Common land". In the case of common land not assigned the type of tenure is owner farming.

We don't use a specific questionnaire to collect information on common land.

4.2 Possible problems encountered in relation to the collection of information on common land and possible solutions for future FSS surveys
No problems have been faced in the collection of this information. 
4.3 Total area of common land in the reference year
Common lands are part of the list register or framework created before data collection. This means that basically only common lands holdings with UAA less than an hectare are excluded.
Common land not assigned/leased during the crop production year of the survey totalled 3 234 232 hectares and 1 433 047 hectares of UAA. 
4.4 Number of agricultural holdings making use of the common land or Number of (especially created) common land holdings in the reference year
3 370 common land units have been included in FSS.
940 common land units have been included in the sample for FSS. 


5. Differences across regions within the country
No one 


6. Organic farming. Possible differences between national standards and rules for certification of organic products and the ones set out in Council Regulation No.834/2007
Spanish legislation has been adopted from the European legislation.
8.1.1. Asymmetry for mirror flow statistics - coefficient

[Not requested]

8.2. Comparability - over time
1. Possible changes of the definition of the agricultural holding
There have not been changes since last survey. 


2. Possible changes in the coverage of holdings for which records are sent to Eurostat
Only the threshold for Christmas trees was removed. "Christmas trees" is no longer a characteristic in 2016 FSS. The impact is seen as negligible.

Instead of B_1_7_1$ha+B_1_8_1$ha+B_4_1$ha+B_4_2$ha+B_4_5$ha+B_4_6_1$ha+B_4_7$ha >= 0.2,

we now have B_1_7_1$ha+B_1_8_1$ha+B_4_1$ha+B_4_2$ha+B_4_5$ha+B_4_7$ha >= 0.2.


3. Changes of definitions and/or reference time and/or measurements of characteristics
Changes to the earlier surveys and censuses are due to changes in Community Regulations on the characteristics and definitions to be used and the practical adaptation of the Spanish questionnaire to European requirements without additional questions of national interest, except for the distinction between dry and irrigated surfaces for each crop.

There are no differences with FSS 2013.


4. Changes over time in the results as compared to previous FSS, which may be attributed to sampling variability
Those characteristics not included in the precision requirements (Annex IV of the Regulation 1166/2008) with a low prevalence could have changes since last survey due to sampling variability. 


5.Common land
5.1 Possible changes in the decision or in the methodology to collect common land
Common land was treated in exactly the same way as in previous censuses and surveys. 
5.2 Change of the total area of common land and of the number of agricultural holdings making use of the common land / number of common land holdings
All changes of the total area and the number of common land units are due to the different assignment of the land between holdings (See 8.1. Comparability - geographical - item 4.1).

The total area of common land has decreased from 3 511 487 ha in 2013 to 3 234 232 ha in 2016 and UAA also has decreased from 1 605 369 ha to   1 433 047 ha in the same period.

The common land units have decreased from 3 803 in 2013 to 3 370 in 2016.


6. Major trends on the main characteristics compared with the previous FSS survey
Main characteristic Current FSS survey Previous FSS survey Difference in % Comments
Number of holdings 945 024 965 002  -2.07%   
Utilised agricultural area (ha)  23 229 753 23 300 221  -0.30%   
Arable land (ha) 11 462 915 11 294 620  1.49%   
Cereals (ha) 6 610 831 6 408 848 3.15%   
Industrial plants (ha) 932 418 947 191  -1.56%   
Plants harvested green (ha) 797 399 794 988  0.30%   
Fallow land (ha) 2 388 077 2 423 435  -1.45%   
Permanent grassland (ha) 7 615 991 7 962 038  -4.35%   
Permanent crops (ha) 4 149 724 4 042 357  2.66%   
Livestock units (LSU) 14 442 532  14 501 694  -0.41%   
Cattle (heads)  6 090 591 5 776 381  5.43%   
Sheep (heads) 15 862 164 15 952 621 -0.57%   
Goats (heads) 2 490 681 2 391 484 4.15%   
Pigs (heads) 23 946 459 24 166 539 -0.91%   
Poultry (heads) 203 105 802 205 823 079 -1.32%   
Family labour force (persons) 1 515 009  1 437 191 5.41%   
Family labour force (AWU) 472 387  485 961  -2.79%   
Non family labour force regularly employed (persons) 351 199  345 495  1.65%   
Non family labour force regularly employed (AWU) 186 497 175 092 6.51%   
8.2.1. Length of comparable time series

[Not requested]

8.3. Coherence - cross domain
1. Coherence at micro level with other data collections
The results were evaluated continuously throughout editing. 

During centralised editing, the application indicated whether the holding was included in another source of information, such as the Agricultural Census 2009, FSS 2013, IACS or the Livestock Register, among others. This allowed the editor to compare information at micro level.


2. Coherence at macro level with other data collections
After automatic imputation and its data analysis, the aggregate variables were calculated: Annual Working Units (AWU), Livestock Units (LSU), Total Standard Output (TSO) and Farm Type (FT). The latter two were obtained after cross-referencing our dataset with the Standard Outputs (SO) dataset, provided by the Ministry of Agriculture.

Prior to final approval of the data, the results were again compared. The results were checked with other data sources before their final approval: Agricultural Census 2009, Farm Structure Surveys, Yearbook of the Ministry of Agriculture and Rural Property Register. 
In the comparisons, we had into account that it could be differences due to several reasons:

- In FSS, the unit of information is the farm above the mentioned threshold in the Regulation 1166/2008 and the information units in crop statistics are parcels based in an area frame sampling design.
- The reference periods are also different. 
- The used definitions could be not the same. 
- Due to the fact that FSS excludes small holdings, in general, FSS data are smaller than crop statistics data. The differences between these two operations could be greater in those cultures that have small surfaces per holding like vineyards, olives, and other cultures depending on the Region.

Regarding livestock characteristics, the differences could also be explained by several reasons. 
- The periods of reference of both statistics are not the same. 
- The date of reference in FSS is 30th September of 2016. The population investigated in FSS includes all farms with 1 or more LSU and TSO equal to or above 900 Euros. 
- Our figures are related to number of heads and not places.

The FSS results are coherent with those collected by the Ministry of Agriculture but they could not be combined as both operations have different units of observation.

8.4. Coherence - sub annual and annual statistics

[Not requested]

8.5. Coherence - National Accounts

[Not requested]

8.6. Coherence - internal

[Not requested]

9. Accessibility and clarity Top
9.1. Dissemination format - News release

[Not requested]

9.2. Dissemination format - Publications
1. The nature of publications
2016 FSS results contain general information about land use, land tenure, holding size, legal status, livestock, organic production, irrigation methods employed and source of irrigation water used, rural development, soil and manure management practices applied, labour force and other gainful activities of the holdings. The results are disaggregated at Autonomous community and national level and classified by UAA and Farm Type. 


2. Date of issuing (actual or planned)
2016 FSS results were published on the INE website on 15th of December of 2016. 


3. References for on-line publications
All results could be accessed in the following link:
9.3. Dissemination format - online database
Dissemination format - online database
The dissemination of the survey is made by tabulations which are published in Ine website (See point 9.2 Dissemination format - Publications)

Microdata will be published in the same link as tabulations.

9.3.1. Data tables - consultations
Data tables - consultations
Not available
9.4. Dissemination format - microdata access
Dissemination format - microdata access
In addition to the planned publications, the micro datasets will be available to all users. These datasets have been anonymised by removing all variables identifying the holder. 
9.5. Dissemination format - other

[Not requested]

9.6. Documentation on methodology
1. Available documentation on methodology
The methodology of the FSS 2016 was published in early 2016, prior to completion of the fieldwork. This publication details the background, aims, content, concepts and definitions, units of measurement and types of holding, data collection and dissemination of data to be used.

This methodology can be accessed through the same link as tabulations and microdata.


2. Main scientific references
- Bethel: Répartition de l'echantillon dans les enquêtes à plusieurs variables. Techniques d'enquêtes, 1989, vol.1, pages 49-60.

- Cochran: Sampling Techniques, Mexico, 1980 pp. 169-174.

- Julien and Maranda, Le plan de sondage de l'enquête nationale sur les fermes de 1988. (Techniques d'enquête 1990, vol. 16 nº 1)

9.7. Quality management - documentation
Quality management - documentation
Relative standard errors (RSE) will be published in INE website with tabulations, microdata and methodology used. 
9.7.1. Metadata completeness - rate

[Not requested]

9.7.2. Metadata - consultations

[Not requested]

10. Cost and Burden Top
Co-ordination with other surveys: burden on respondents
The sample of FSS 2016 has not been coordinated with any survey.

11. Confidentiality Top
11.1. Confidentiality - policy
Confidentiality - policy
The statistical data provided to the National Statistics Institute is protected by statistical secrecy. Statistical Secrecy is a guarantee and trust mechanism for respondents that implies the protection of the data that is obtained for statistical purposes.

Chapter III of the aforementioned LFEP regulates all aspects of statistical secrecy. According to its content:

  • The personal information obtained by the statistical services will be the object of protection and will be covered by statistical secrecy.
  • All statistical personnel have the responsibility of maintaining statistical secrecy. The obligation of maintaining statistical secrecy will remain after those persons obligated to maintain it end their professional activities or their link to the statistical services.

In addition to the expected publications, the microdata files are made available to all users. These files are anonymised, deleting all variables that identify the holder.

11.2. Confidentiality - data treatment
Confidentiality - data treatment
The microdata exclude all personal references which permit the identification of the holders. No other confidentiality treatment has been applied. 

12. Comment Top
1. Possible improvements in the future
To improve the IRIA application with the acquired experience in this survey and another future surveys performed by INE. 


2. Other annexes
Not available.

Related metadata Top

Annexes Top