Farm structure (ef)

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

Compiling agency: Central Statistics Office

Time Dimension: 2016-A0

Data Provider: IE1

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

Download


1. Contact Top
1.1. Contact organisation
Central Statistics Office
1.2. Contact organisation unit
Agriculture Surveys Section
1.5. Contact mail address
Central Statistics Office, Skehard Road, Cork, Ireland


2. Statistical presentation Top
2.1. Data description
1. Brief history of the national survey 

A Farm Structure Survey (FSS) is carried out between Censuses to measure changes in Farm Structure. The first Census of Agriculture in Ireland was carried out in 1847, and annually thereafter until 1953. Between 1960 and 1980 Censuses were carried out at 5 yearly intervals. From 1980 Censuses were carried out at 10 yearly intervals.

Farm Structure Surveys were carried out on three occasions in the 1980s ('83, '85, and '87), 1990s ('93, '95, and '97) and 2000s ('03, '05, '07). There has also been one subsequent Farm Structure Survey in 2013 since the 2010 Census of Agriculture. 

The data for the 2013 Farm Structure Survey was collected using an 8-page postal questionnaire. This questionnaire data was, for the first time for an FSS, supplemented by the use of administrative data from the Irish Ministry of Agriculture. The administrative databases utilised were the IACS and Bovine Registers. The use of these administrative databases has continued in subsequent years including Farm Structure Survey 2016. The 2010 Census of Agriculture was the first instance  that utilised these administrative databases. 

The statistical register compiled for the 2010 Census of Agriculture was updated for FSS 2013 and FSS 2016 by adding new administrative records of agriculture holdings or livestock herds created since the 2010 Census. These administrative records were provided by the Irish Ministry of Agriculture, known as the Department of Food, Agriculture and the Marine (DAFM).  New administrative ‘births’ were added to the existing register of 139 860 holdings from the 2010 Census giving a total sample frame of 144 931 holdings while also accounting for deaths.

Questionnaires were issued to 55 909 holdings in the week preceding the reference date of 1st June 2016. Three reminders were issued at approximately fortnightly intervals to maximise the response rate.

 

2. Legal framework of the national survey 
- the national legal framework

The statistical activities of the CSO are governed by the Statistics Act, 1993.
See: http://www.irishstatutebook.ie/1993/en/act/pub/0021/print.html

This act provides the legislative framework for the CSO. It sets out the right of the Office to conduct statistical inquires (See Part III - Collection of Information of the Act above)

- the obligations of the respondents with respect to the survey Part III (Section 24) of the Statistics Act specifies that a person may be invited to provide information on a voluntary basis. As the level of voluntary response to agricultural surveys in Ireland was already deemed satisfactory, a ministerial order was not deemed necessary for FSS 2016.
- the identification, protection and obligations of survey enumerators Part II (Section 20, 21 and 22) of the Statistics Act specifies the responsibilities of Central Statistics Office staff. It should be noted that as the Farm Structure Survey is a postal survey,  enumerators do not directly visit farms/respondents.
2.2. Classification system

[Not requested]

2.3. Coverage - sector

[Not requested]

2.4. Statistical concepts and definitions
List of abbreviations

AIM - Animal, Identification and Movement System

CCS - Corporate Client System

CSO - Central Statistics Office

DAFM - Department of Agriculture, Food and the Marine

DMS - Data Management System

IACS - Integrated Administration and Control System

NSI - National Statistics Institute

OCI - Scanning System Utilised

SPS - Single Payment Scheme

 

2.5. Statistical unit
The national definition of the agricultural holding

An agricultural holding was defined, in line with the definition in Article 2 of Council Regulation 1166/2008 as ‘a single unit, both technically and economically, which has a single management and which undertakes agricultural activities within the economic territory of the European Union, either as its primary or secondary activity’ 

Activities considered 'agricultural' for the purposes of the definition above as outlined in Annex I of Regulation 1166/2008, include the growing of perennial and non-perennial crops, plant propogation, animal production, mixed farming and/or those maintaining agricultural land in good agricultural and environmental condition (under 01.61 of NACE Rev. 2).

2.6. Statistical population
1. The number of holdings forming the entire universe of agricultural holdings in the country
The NSI Agricultural Register used contained 144 931 agricultural holdings.

 

2. The national survey coverage: the thresholds applied in the national survey and the geographical coverage
A minimum size threshold of 1 hectare, as per Article 3 of Regulation 1166/2008, was not applied in advance of the FSS2016 survey. Excluding such units could significantly reduce state-level results for some characteristics that are more frequent on smaller holdings (e.g. goats, pigs, or poultry). In addition, there is a sizeable amount of common land in the state, which can allow very small land holders to keep significant numbers of livestock (sheep in particular). The target population of FSS 2016 was all agricultural holdings in Ireland, irrespective of size.

 

3. The number of holdings in the national survey coverage 
The estimated number of agricultural holdings in 2016 was 137 560, which includes 27 commonage units.

 

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

 

5. The number of holdings in the population covered by the records transferred to Eurostat
All 137 533 holdings were covered by the records transferred to Eurostat, along with  27 commonage areas. Therefore, the file transmitted contained 137 560 observations.

 

6. Holdings with standard output equal to zero included in the records sent to Eurostat
SO was not provided in micro-data file.

 

7. Proofs that the requirements stipulated in art. 3.2 the Regulation 1166/2008 are met in the data transmitted to Eurostat
No threshold was used in the FSS and the full population of holdings on the agricultural register was taken into consideration for FSS 2016. There is no threshold for inclusion on the register.

 

8. Proofs that the requirements stipulated in art. 3.3 the Regulation 1166/2008 are met in the data transmitted to Eurostat
No threshold was used in the FSS and the full population of holdings on the agricultural register was taken into consideration for FSS 2016. There is no threshold for inclusion on the register.
2.7. Reference area
Location of the holding. The criteria used to determine the NUTS3 region of the holding
The majority of the total area of the holding where the holding is located.The largest parcel of land on each farm was used as the reference parcel to allocate a farm into the various NUTS regions.
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)

For survey data, a reference date for the land characteristics was June 1st 2016, with the reasonable assumption that the crop in the ground on that date was the main crop for the 12 month period ending June 1st 2016. The administrative data source used a reference date of 31st May 2016, with a similar assumption.

The reference date for the livestock characteristics was June 1st 2016 in the survey. This is also the reference date for bovines in the administrative data.

The reference period for the all labour force characteristics was the twelve-month period prior to June 1st 2016.

The reference period for the rural development measures was the three-year period between January 1st 2014 and December 31st 2016.

2.9. Base period

[Not requested]


3. Statistical processing Top
1.Survey process and timetable
1. FSS 2016 Project Initiation
  • December 2015: Transmission of NE/NS characteristics to ESTAT.
  • February 2016: Project Plan development
  • February - March 2016: Consultation with Stakeholders

2. Census Preparations/Survey Design

  • Feb-Mar 2016: Design of questionnaire
  • March 2016: Testing of questionnaire for electronic scanning
  • May 2016: Printing of questionnaire
  • Apr-May 2016: Selection & Training of temporary processing staff

A pilot survey was not carried out as the FSS questionnaire was quite similar in content and design to questionnaires issued in previous agricultural surveys. The questionnaire was circulated to a small number of people with agricultural expertise for review prior to being launched and this process provided constructive feedback which was incorporated into the final version of the questionnaire. The accompanying information booklet was designed to add clarity to the data being collected.

3. Survey Management

  • Feb 2016 - April 2016: Update Farm Register
  • May 2016: Issue FSS2016 postal Questionnaire
  • June-Sept 2016: Returns receipted
  • June 2016: 1st reminder notice issued
  • July 2016: 2nd reminder notice issued
  • July 2016: 3rd reminder notice issued

4. Data Capture/Data Processing /Data analysis

  • June-Oct 2016: Returns electronically scanned and verified
  • August 2016-Dec 2016: Data on paper returns edited
  • Jan 2017 - May 2017: Integrate provisional administrative data with survey returns
  • May 2017 - Oct 2017 : Integrate final administrative data with survey returns and check data for coherence
  • Jan 2017 - June 2017: Imputation & weighting; check aggregates
  • Jan 2017- Feb 2017: Impute for missing units to provide a full Census for FSS2016.

5. Data Dissemination

  • September 2016: Pig results (2016) published
  • September 2016: Provisional crops and livestock results published (state-level only)
  • March 2017: Final crops and livestock results published (regional level)
  • December 2017: First FSS2016 Micro-data file coded to Eurofarm format & transmitted
  • February 2018: Draft National Methodological report  transmitted;
  • February 2018: Final FSS2016 Micro-data file coded to Eurofarm format & transmitted;
  • July 2018: Final validated FSS2016 Micro-data;
  • November 2018: Final FSS2016 Micro-data including rural development measures. 

 

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

The FSS2016 was carried out by the Agriculture Division within the Central Statistics Office.

The project team consisted of 8 core processing staff, 2 Statisticians (one of whom was Project Manager) and 1 Senior Statistician. A further 10 CSO staff were assigned on a temporary basis for data processing between May and September 2016. Together these staff made up the project team.

This team was responsible for all aspects of FSS2016 from project planning and survey management through to dissemination of results. The team updated the farm register, designed and issued the FSS questionnaire and reminder notices, receipted and scrutinised questionnaires returned, scanned and verified data. The team then moved on to merging questionnaire data with administrative data and finally analysis of data and preparation of results for transmission to Eurostat and national publications.

It was also responsible for a specialist survey of large pig units (330 units), carried out as part of the FSS.

The CSO has its own internal Printing section which produced the questionnaires and accompanying information booklets. The CSO has an Office Services Unit which provided support in managing the large volumes of outgoing and incoming post.

 

3. Serious deviations from the established timetable (if any)
Not available.
3.1. Source data
1. Source of data

The agricultural register used for FSS2016 was based on the register used for FSS2013 with new farm holdings ('births') instituted since 2010. The NSI Agriculture register contained 144 931 agricultural holdings. By combining both sample data and administrative data and imputing for any further variables required, a Census has been created for FSS 2016 for most of the characteristics. All of the main characteristics are provided for the full population through a combination of administrative and survey data sources.

FSS 2016 provides a farm population of 137 560 farms as follows:

i) 37 007 records from an FSS2016 sample survey PLUS

(ii) 100 001 records confirmed active on administrative databases held by Ministry of Agriculture  PLUS

(iii) 108 pig farms not in (i) or (ii) above but confirmed active in a specialised Pig Survey held on same reference date  PLUS

(iv) 417 records not in (i) or (ii) or (iii) but confirmed active in FSS 2016 and no evidence in the interim to suggest activity has ceased PLUS

(v) 27 commonage areas.

FSS 2016 used a combination of both administrative records and completed paper questionnaires and imputation techniques to compile the required data. In an effort to reduce the response burden on farmers, all questions relating to organic farming, cattle, cereals, potatoes and most crops were eliminated from the FSS 2016 questionnaire as the relevant data were available from administrative data files provided by DAFM. 

A supplementary pig questionnaire was sent to a sample of 329 pig farmers. These were farmers who returned 100 or more pigs in one or more of the preceding four pig surveys (June 2014, December 2014, June 2015 and December 2015).  This questionnaire allowed the pig producer to provide additional breakdowns of the breeding sows and of the non-breeding categories, a format which they were already familiar with. 

 

2. (Sampling) frame

The Farm Register was compiled using a combination of the pre-existing CSO Farm Register and two administrative databases held by the DAFM, namely the Corporate Client System (CCS) and the Animal Identification and Movement (AIM) system:

(i) The pre-existing CSO Farm Register was created for the last FSS which took place in 2013. This register was maintained by the CSO Agriculture Register Section and updated with births and deaths identified in the annual June and December surveys between 2013 and 2016 which incorporates BPS (IACS) data. It was used as the sampling frame for every agriculture survey that was carried out by CSO since 2013. The availability of administrative files since 2010 ensures that now all entries on this register can be checked on an annual basis for activity and accurate contact details. The CSO register holds only contact information and location details. The register does not hold any structural variables.

(ii) The CCS database was received from DAFM in Spring 2016. This contained  records consisting of the name, address, telephone number, email, date of birth, and herd number of every farmer considered to be active by DAFM. The CCS database is separate to the IACS database but contains all of the holdings that are on the IACS system. The CCS database was used solely for the purposes of building the register. No statistical data was extracted from the CCS. 

(iii) The AIM database was received from DAFM in Spring 2016. Any record without a corresponding entry in CCS (ii above) were added to the Register. 

The resulting register was used as the frame for the FSS2016.

Type of frame is multiple list frames.

 

3. Sampling design
3.1 The sampling design

The sample was a stratified one-stage sample of holdings (probability design).

3.2 The stratification variables

The selection process was based on 53 initial strata. There were a number of strata for whom we wanted to have a 100% sample. Farms selected for strata 1 and 2 were defined due to their specialist nature and their relatively small population. Sheep farms were selected in stratum 3, as we do not have an administrative data source for sheep and a matched sample is required annually for which we have a minimum sample size. Stratum 4 included all new births. The remainder of the sample was selected in strata 5-53.

The strata are:

  1. All Poultry farms (farms with greater than or equal to 100 birds),
  2. All Mushroom farms (farms with mushrooms in FSS2013)
  3. Sheep farms, we generated a matched sample with the June 2015 survey, which resulted in 5 426 sheep farms.
  4. Stratum 4 consisted of all new births according to administrative files. Once the births were checked and duplicates identified, the final number of births was 5 273.
  5. Strata 1-4 initially resulted in 11 229 units selected into the sample. In order to generate a final sample of approximately 56 000, 45 000 units needed to be selected from the remaining 96 409 units on the sampling frame that were not included in strata 1-4 . These units were selected based on NUTS3 region and farm size and resulted in 49 additional strata.
3.3 The full coverage strata
The following farms had complete enumeration:
  1. All Poultry farms (farms with greater than or equal to 100 birds),
  2. All Mushroom farms (farms with mushrooms).

as well as the new births.

3.4 The method for the determination of the overall sample size
The final sample was checked for duplicates or inactivity on the Agricultural Register, which resulted in dropping 281 units from the sample. This resulted in a final sample of 55 909.
3.5 The method for the allocation of the overall sample size

In stratum 3, firstly, we generated a matched sample with the June 2015 survey. We selected all farms in stratum 1, stratum 2 and stratum 4. 

In strata 5-53, we applied the Neyman allocation method according to farm size and NUTS3. Area farmed was the variable used to calculate variability with the 49 strata.

3.6 Sampling across time
A new sample was selected from the register.
3.7 The software tool used in the sample selection
The sample was selected in SAS.
3.8 Other relevant information, if any
Not available.

 

4. Use of administrative data sources
4.1 Name, time reference and updating

The main administrative data files used were:

  1. the Integrated Administration and Control System (IACS). This is provided according to Council Regulation (EC) No 1782/2003 of 29 September 2003 establishing common rules for direct support schemes under the common agricultural policy and establishing certain support schemes for farmers and amending Regulations (EEC) No 2019/93, (EC) No 1452/2001, (EC) No 1453/2001, (EC) No 1454/2001, (EC) 1868/94, (EC) No 1251/1999, (EC) No 1254/1999, (EC) No 1673/2000, (EEC) No 2358/71 and (EC) No 2529/2001[1].
  2. the System for the Identification and Registration of Bovine Animals (AIM system). This is  provided according to Regulation (EC) No 1760/2000 of the European Parliament and of the Council of 17 July 2000 establishing a system for the identification and registration of bovine animals and regarding the labelling of beef and beef products and repealing Council Regulation (EC) No 820/97[2].

Data on crops, cereals and potatoes were obtained from DAFM’s Single Payment Scheme (Council Regulation No 1782/2003) while all data on cattle was obtained from DAFM’s Animal Identification and Movement system (Council Regulation No 1760/2000). The use of these administrative data sources was provided for under Article 4.1 of the FSS regulation. The earliest application date for the SPS is March, and the closing date for applications was May, so while the reference date does not correspond exactly to our FSS reference date of 01/06/2016, the information in the SPS dataset was still usable for FSS2016 as it represented what was in the ground for the harvest year.

In addition to the above, the Corporate Client System database was received from our Ministry of Agriculture (DAFM) in Spring 2016 for the purpose of enhancing the Agriculture Register held by this NSI. This contained records consisting of the name, address, telephone number, email, date of birth, and herd number of every farmer considered to be active by DAFM. The CCS database is held separately to the IACS database but does contain all of the holdings that are on the IACS system. The CCS database was used solely for the purposes of building the register. No survey data was extracted from the CCS.

[1] OJ No. L270, 21.10.2003,p.1.

[2] OJ No. L204, 11.08.2000,p.1.

4.2 Organisational setting on the use of administrative sources
The Statistics Act, 1993 (Part IV) grants the CSO right of access to records of public authorities for statistical purposes (with a number of exceptions). Specifically under the Act, the CSO may request any public authority to consult and co-operate with (the CSO) for the purpose of assessing the potential of the records of the authority as a source of statistical information and, where appropriate and practicable, developing its recording methods and systems for statistical purposes. This underpins co-operation with the Department of Agriculture on the subject of its farm registers. (See Part IV - Use of Records of Public Authorities for Statistical Purposes of the Act above.)
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 reporting unit in the SPS (IACS) and the AIM databases is the Herd Number. Records were joined across databases using the Herd Number identifier.

IACS: Data for each parcel of land in the IACS is aggregated to the farm level.

In some but not all instances, farm holdings may have more than one herd number which complicates the linking process. Multiple herd numbers were combined onto one record where more than one herd number belonged to a holding.
- coherence of definitions of characteristics Bovine Register:
All C_2 to C_2_99 characteristics can be directly computed based on the age and gender of the animal as recorded in the AIM database.
   
- coverage:      
 over-coverage under-coverage misclassification   No over-coverage.  

IACS: The percentage of farmers who do not apply for payment under this scheme is now very small in number and in land area.

 

 

A small number of small farms without Bovines and without IACS payment entitlements, may fall outside of the scope of administrative databases. We estimate this to be 1.3% of farms ((1779+60)/137560).

The small farms are captured through the NSI Agriculture Register. To reduce the possibility of under-coverage, all of the administrative databases were merged together with the existing CSO agriculture register and the union of all of these records was used to establish the 2016 Register of Farm holdings.
Bovine Register: The fact that the AIM system is the definitive register of cattle owners in Ireland implies that the coverage provided by this data source is more complete than by survey.
Farms are not classified on administrative databases.    
 multiple listings Names and addresses and dates of birth examined.    
- missing data All cells were complete in the administrative database.

 

 
- errors in data

Referred back to Ministry of Agriculture. Very very rare.

Bovine Register: The fact that the AIM system is the definitive register of cattle owners in Ireland implies that the system of fines for non-compliance with the AIM scheme incentivises farmers to keep its information up to date. As farmers must supply the date of birth when registering an animal, the AIM data for age is likely to be much more accurate than for survey data.

   
- processing errors Well-established processing systems which are carried out bi-annually (IACS) and Quarterly (Bovine Register) and also well documented.    
- comparability   Not available.  
- other (if any)   Not available.  

 

4.5 Management of metadata
Metadata on the administrative data sources used are stored electronically in a series of production manuals. These manuals are text documents and are stored and maintained on the Agriculture server. 
4.6 Reporting units and matching procedures
The reporting unit in the SPS (IACS) and the AIM databases is the Herd Number. Each herd number represents a farm holding in Ireland. All holdings are identified by their herd number and all data is linked based on herd number as this is the unique identifier. It is important to point out that all holdings have a herd number, irrespective of whether they hold any livestock or not.
4.7 Difficulties using additional administrative sources not currently used
Nothing to report.
3.2. Frequency of data collection
Frequency of data collection
A Farm Structure Survey (FSS) is carried out approximately every three years to measure changes in farm structure between Censuses. Between Census 2010 and Census 2020 there are Farm Structure Surveys in 2013 and 2016.
3.3. Data collection
1. Data collection modes

By combining both sample survey data and administrative data and imputing for any further variables required, a Census has been created for FSS2016 for most of the characteristics.

The survey data was collected as follows:

An 8-page A4 sized questionnaire (see Annex 3.3-5. FSS 2016 Questionnaire) was issued to all farm holdings in the week prior to June 1st 2016 to be completed and returned to the CSO by Tuesday 8th June 2016. This was accompanied by an information booklet (see Annex 3.3-5. FSS 2016 Information Booklet) with detailed notes on each section of the questionnaire. Three reminders were issued in order to maximise the response rate.

A separate 2-page A4 sized questionnaire (see Annex 3.3-5. Pig Survey Form 2016) was also issued to all specialist pig-producers.

The paper questionnaires returned to the CSO were batched, receipted and scrutinised. They were then electronically scanned, verified and edited. 

The FSS survey data were collected entirely by post (i.e. no interviewers). Each questionnaire issued included a pre-addressed freepost reply envelope. The return address on the reply envelope was to some dedicated Post Office Boxes which were used exclusively for the Census of Agriculture. The advantage of this method was that post was segregated from other post when it was delivered to the CSO and the national Post Office was able to provide an exact count of how many envelopes were returned this way. The envelopes were mechanically cut open across the top and the questionnaire held within was removed manually.

 

2. Data entry modes
OCI scanning directly captured the respondents' reply to every question on the form. Each question had an answer field with a specific box for each digit to be entered (or ticks in the case of tick-box questions). The OCI software created an image of each questionnaire. Where the OCI software could not clearly identify a numeric or character value entered on the questionnaire, the software 'held' the questionnaire for review by a member of staff. Once this verification process was completed the data on the questionnaire was written to a flat file for import into the CSO Data Management Systems' Data Entry module. A series of checks were run to ensure that the data returned was consistent (e.g. to ensure that totals equaled the sum of their parts, to ensure that livestock holdings returned sufficient grazing land etc.). Once these checks were completed, a 'clean-unit' copy of the data was available to be merged with administrative data and this produced a complete, or almost complete, record for each holding.

 

3. Measures taken to increase response rates

The rate of return of questionnaires was monitored daily.  When the response rate began to decline below targets, a reminder notice was issued to those who had not yet responded. In total three reminder notices were issued between June 2016 and August 2016. The second and third reminder notices also included a copy of the questionnaire, in case the original had been mislaid. A limited number of non-respondents were phoned.

The questionnaire was issued to 55 909 holdings in the last week of May 2016. To increase the response rate three written reminders were posted out to non-respondents between June-August 2016. 73.1% (40 847) of the issued questionnaires were returned.  Of these, 37 007 were considered active farms from the information on the questionnaire.  

 

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

144 931 
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

55 909
3 Number of ineligible holdings 1 636
3.1

Number of ineligible holdings with ceased activities

This item is a subset of 3.

1 636
4

Number of holdings with unknown eligibility status

4>4.1+4.2

0
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=5.1+5.2

54 273
5.1

Number of eligible non-responding holdings

5.1>=5.1.1+5.1.2

17 266
5.1.1 Number of eligible non-responding holdings – re-weighted
5.1.2 Number of eligible non-responding holdings – imputed 17 266
5.2 Number of eligible responding holdings 37 007
6

Number of the records in the dataset 

6=5.2+5.1.2+4.2

54 273

(eligible returns from sample)

 * Figures in this table exclude Common land units. 

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


Annexes:
3.3-5. FSS 2016 Questionnaire
3.3-5. FSS 2016 Information Booklet
3.3-5. Pig Survey Form 2016
3.4. Data validation
Data validation

The data undergoes checks and controls throughout the processing stages. This takes place across three stages: scrutiny, verification and editing.

Scrutiny is a manual process undertaken by processing staff. The function of scrutiny is to detect and rectify errors in advance of data entry. It involves visually examining the questionnaire, page by-page. Where a ‘yes/no’ tick box was left unchecked or an inconsistency in the data was encountered, it was manually amended on the questionnaire where possible.

After data entry, the data enters verification stage. This is done electronically. Illegible digits and data are highlighted and corrected on screen by referring to previous returns or by deduction.  The data then enters the edit phase where data is passed through a range of pre-programmed edit checks. Here, arithmetic checks, range checks and consistency checks (with previous returns) are carried out and data is examined where these checks fail and then edited if necessary.

Data was processed through our Data Management System (DMS), which contains certain edit rules which ensure some basic unit level consistency. Edit rule failures were examined and corrected. Data was then processed using SAS for further edit checks.

Each of the three stages of the data validation was carried out by the processing staff in the CSO made up of temporary staff and more experienced supervisory staff.

3.5. Data compilation
Methodology for determination of weights (extrapolation factors)
1. Design weights
The design weights were the inverse of the sampling probability within the various strata. Stratum weight h= Nh/nh, where Nh= number of farms in stratum h and nh= number of sample farms in stratum h.
2. Adjustment of weights for non-response
To account for non-response, the design weights were adjusted so that, Stratum weight h= Nh/(nh-mh), where Nh= number of farms in stratum h, nh= number of sample farms in stratum h, and mh= number of non-respondent sample farms in stratum h.
3. Adjustment of weights to external data sources
A small overall adjustment factor (137560/139595=.9854) was applied to the weights to account for the very slight change in population betweeen 2013 and 2016.
4. Any other applied adjustment of weights
 Not applicable
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

not applicable

5.1. Relevance - User Needs
Main groups of characteristics surveyed only for national purposes 

The main groups of national characteristics surveyed are decided based on EU FSS legislation, i.e. Regulation 1166/2008.

A small number of additional variables were collected for national purposes. The need for these variables was identified through a consultation process with the main stakeholders prior to the survey design stage. Specifically these variables related to:

  • Sheep: A  more detailed breakdown of sheep, consisting of rams, ewes (both under and over 2 years) and other sheep (both under and over 1 year).
  • Poultry: The sub-division of both broilers and turkeys into breeding birds and table birds. 
  • Equidae: The sub-division of equidae into thoroughbred, other horses and mules, jennets and asses.
  • Deer: Number of farmed deer.
  • Administrative Burden: the number of minutes taken to complete the FSS2016 questionnaire.
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 find 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

not applicable

6.1. Accuracy - overall
Main sources of error

The following are considered to be the main sources of error in FSS2016:

  • coverage error,
  • non-response error,
  • measurement error.
6.2. Sampling error
Method used for estimation of relative standard errors (RSEs)
See Annex.


Annexes:
6.2. FSS 2016 variance estimation
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

There are no cases where the estimated RSEs are above the thresholds where precision requirments are applicable. Crop and livestock characteristics are collected for each holding in the population.

The RSEs have been however calculated as if crop and livestock characteristics were collected based on the sample, in order to illustrate the quality of the sample for the characteristics actually collected only for the sample.



Annexes:
6.2.1-1 Relative standard errors
6.3. Non-sampling error

 not applicable

6.3.1. Coverage error
1. Under-coverage errors
All necessary steps are taken to ensure full coverage of the population. The Agriculture Register, finalised after FSS 2013, was further updated in April 2016 (prior to FSS2016) to add 5 273 new 'births' which had been identified as newly-active holdings on Ministry of Agriculture's administrative databases. Therefore, the Agriculture register contained 144 931 entries and was considered to be very comprehensive. The only units that could have been excluded were those farming but not registered on either of the two administrative databases (IACS & Bovine Register). However, the likelihood of a new farm not falling into one of these two databases is considered low.

  

2. Over-coverage errors
While 5 273 'births' were added to the register, it is not always easy to identify farm 'deaths'. However, page 1 of the FSS questionnaire asks the respondent to indicate if the holding has been sold or leased or if the registered holder has retired or is deceased. These units are subsequently marked as inactive and considered 'out-of-scope'. These out-of-scope units are taken into consideration when calculating survey weights, in that only in-scope responses are included when calculating the non-response weight.
2.1 Multiple listings 
While 5 273 'births' were added to the register, it was difficult to identify farm 'deaths' and this may lead to duplicate entries on the register if the new record related to a new owner took over an existing farm holding. While every effort is made to eliminate these prior to issuing questionnaires, it was possible that some farms received two questionnaires. In some of these cases, farmers returned the second (blank) questionnaire with the completed questionnaire. To eliminate cases where two questionnaires were completed for the same holding, a thorough examination of data was carried out to identify records with identical data. This was done primarily using name and address matching, but also using several of the key variables. In all, approximately 383 duplicates were identified. These were considered inactive and out of scope and excluded from the non-response weight calculation.

 

3. Misclassification errors
Units were initially classified according to data collected in FSS 2016  which was very comprehensive, extensively validated and agreed with Eurostat. Therefore, we do not consider misclassification to have been an issue. There is the possibility that units have increased/decreased in size and economic size in 2016 and moved into a new class size as a result but this is likely to have only affected a very small number of holdings. Therefore, there was no adjustment of strata prior to weighting up.

 

4. Contact errors
The contact data was provided by the Agriculture Register. In some cases, the holder could not be reached at that address and the questionnaire was returned unopened. This occured in just 40  cases. These were considered inactive and out of scope and excluded from the non-response weight calculation.

 

5. Other relevant information, if any
Not available.
6.3.1.1. Over-coverage - rate
Over-coverage - rate
 Estimated at approximately 2.8% at the level of the sample.
6.3.1.2. Common units - proportion

[Not requested]

6.3.2. Measurement error
Characteristics that caused high measurement errors
The  FSS2016 Survey Instrument is a paper questionnaire issued by post and self-completed by the respondent, not by a trained interviewer.  Therefore, the interpretation of certain questions is difficult to control without having a trained interviewer present during completion. However, data are validated extensively at each stage of processing for consistency (with previous responses/ external data sources) and for coherence.  Re-interview/Re-surveying does not occur.
6.3.3. Non response error
1. Unit non-response: reasons, analysis and treatment

Unit non-response occured when a sample unit declined to respond to the questionnaire, despite the issuing of three reminders. Non-response was assumed (as opposed to out-of-scope/inactivity) when a form wasn't returned. Administrative data was utilised where possible for farms which were found to be active on administrative files despite providing no response. Otherwise, imputation was used to impute certain characteristics for the non-sampled units to compile a full census.

However, there was no administrative data or robust imputation method available for a small number of FSS characteristics (other gainful activities, crop rotation and manure management). Therefore, these are available for the responding units only (n=37 007) and as such are weighted variables. Non-response was taken into consideration when calculating weights for these particular variables.

Full non-response was addressed by using administrative data to confirm level of activity and provide data. Therefore, bias due to non-response is considered to have been addressed. The unit non-response rate is 31.8%, non-responding units with unknown eligibility status are treated the same way as the ineligible units. 

 

2. Item non-response: characteristics, reasons and treatment

As all data on bovines and crops were collected from administrative records, only variables collected in the FSS paper questionnaire were affected by item non-response. This seemed to occur mostly in the farm labour, OGA, crop rotation, manure management and training sections. The FSS is a self-completed postal questionnaire (8 pages) and as such there may be respondent fatigue by the time these sections are reached. The data being collected are complex and do not work well in a postal questionnaire with no trained interviewer present during completion. It can therefore be difficult also to determine if the cells are empty due to non-response or are in fact real zero.

Where available, administrative data is used to impute for item non-response or to confirm real zero. In the absence of administrative data , data were imputed using regression if appropriate explanatory variables could be identified.

6.3.3.1. Unit non-response - rate
Unit non-response - rate
The unit non-response rate was 31.8% at the level of the sample.
6.3.3.2. Item non-response - rate
Item non-response - rate
This was not captured.
6.3.4. Processing error
1. Imputation methods
Imputation techniques were used to complete the dataset for the population:

Sheep: The annual Sheep & Goat Census carried out by the Ministry of Agriculture, which provides a register of all sheep producers with a reference date of December of each year. This was used to impute for missing sheep data. The number of breeding females (C_3_1_1) was taken from the Census and an expected non-breeding flock (C_3_1_99) per unit of breeding female was derived controlling for whether the farm was an upland or lowland holding (as this factor influences productivity per breeding female).

Labour: Where the age of the holder was not provided or a unit was not sampled, administrative files were first checked for a date of birth. If this failed, the age at the last FSS in 2013 was checked if available and adjusted accordingly. Finally, if the age could still not be confirmed, the distribution of holder ages across all returns was examined and this distribution was used to randomly assign ages to the missing cases. In returns where the labour force section was left completely blank or in cases where the farm was not directly surveyed in 2016 , regression techniques were utilised to provide a model for labour component of farms based on all available explanatory variables including area farmed, number of livestock, age of holder, gender of holder amongst others. Time spent was also regressed on explanatory variables.

Grass: Where no grassland area was provided for farms with bovines, the number of bovines in each category were used as explanatory variables in predicting a value for area of grassland.

Also, imputation from administrative data or previous surveys was also used to account for unit non-response. 

 

2. Other sources of processing errors

Each form was scrutinised before scanning to highlight any obvious errors. After scanning, the verification procedure ensured that any questionable cells were checked and corrected. 

 

3. Tools used and people/organisations authorised to make corrections
Imputed and regression estimates were compiled by the Statisticians responsible for FSS 2016 and all work was carried out in SAS.
6.3.4.1. Imputation - rate
Imputation - rate

Plants Harvested Green (B_1_9) and Pasture and Meadow excluding rough grazing (B_3 minus B_3_2): Total grassland was available from administrative sources, however the breakdown between Plants harvested green and Pasture and Meadow was unavailable. Therefore the proportional breakdown of pasture and meadow and plants harvested green (from questionnaire) to total grassland (from administrative data) was applied to the non-surveyed units. Imputation rate for B_1_9 and B_3=72.6%

Administrative data was available for all other main characteristics.

6.3.5. Model assumption error

[Not requested]

6.4. Seasonal adjustment

[Not requested]

6.5. Data revision - policy
Data revision - policy
Ireland does not have a revision policy.
6.6. Data revision - practice
Data revision - practice
Some revisions may take place arising from Eurostat validation checks. Otherwise, the data is considered to be final.
6.6.1. Data revision - average size

[Not requested]


7. Timeliness and punctuality Top

not applicable

7.1. Timeliness

not applicable

7.1.1. Time lag - first result
Time lag - first result
FSS2016 results will not be published until the dataset and NMR have been validated and accepted by Eurostat. However, totals for crops and cattles for 2016 gathered from administrative data sources were published in September 2016 as part of annual time series publications for crops & livestock required under EU Regulation 1165/2008 (Livestock) and EU Regulation 543/2009 (Crops).
7.1.2. Time lag - final result
Time lag - final result
FSS2016 results will not be published until the dataset and NMR have been validated and accepted by Eurostat. However, totals for crops and cattles for 2016 gathered from administrative data sources were published in March 2017 as part of annual time series publications for crops & livestock required under EU Regulation 1165/2008 (Livestock) and EU Regulation 543/2009 (Crops). There a three month time lag between the end 2016 (reference) year and the final crops & livestock publication. 
7.2. Punctuality

not applicable

7.2.1. Punctuality - delivery and publication
Punctuality - delivery and publication

Initial Delivery of FSS 2016 to Eurostat = On schedule (Sent Dec 21st 2017)

Final Delivery of FSS 2016 to Eurostat = July 20th 2018

FSS 2016 results will not be published until the dataset and NMR have been validated and accepted by Eurostat.


8. Coherence and comparability Top

not applicable

8.1. Comparability - geographical
1. National vs. EU definition of the agricultural holding
An "agricultural holding" is defined in accordance with Regulation 1166/2008. That is, "a single unit, both technically and economically,which has a single management and which undertakes agricultural activities listed in Annex I within the economic territory of the European Union, either as its primary or secondary activity".

 

2.National survey coverage vs. coverage of the records sent to Eurostat
No thresholds were applied.The population covered in the national survey is the same as that delivered to Eurostat.

 

3. National vs. EU characteristics

The characteristics collected in FSS 2016 correspond to the list of characteristics provided in the Annex of Regulation (EC) No.715/2014. 

The 10th version of the ‘Handbook on implementing the FSS and SAPM definitions’ was used to define the variables.

1 800 hours per annum for a full-time employee is used to calculate the Annual Work Unit.

 

4. Common land
4.1 Current methodology for collecting information on the common land

There are 27 agricultural common land units, covering 427 848 ha. The UAA of the common land is entirely permanent grassland and meadow - rough grazing and the units are recorded at the level of NUTS 4. This estimate was obtained from the Irish agriculture ministry, DAFM. This data is the most accurate data in the state for declared commonage and follows a year-long review undertaken by DAFM of all declared common land in Ireland. This comprehensive review occured in 2009/2010, and consisted of physical inspections of the areas and/or a review of the ortho-imagery. It involved excluding all ineligible features such as scrub, rock, roadways, forests etc to construct accurate areas.

4.2 Possible problems encountered in relation to the collection of information on common land and possible solutions for future FSS surveys
None known.
4.3 Total area of common land in the reference year
The estimate of Common Land was 427 848 hectares for 2016 and was obtained from the Agriculture Ministry, DAFM.
4.4 Number of agricultural holdings making use of the common land or Number of (especially created) common land holdings in the reference year
 27 specially created units.

 

5. Differences across regions within the country
Not applicable for FSS 2016.

 

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
In Ireland, all the EU standards and rules of organic certification as per the relevant Council Regulations viz: Council Reg 834/2007 and 889/2008 have been implemented. However, our Ministry of Agriculture confirmed that higher standards are adopted by two certifying bodies which in some areas  go beyond the EU Regulations.
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
No changes.

 

2. Possible changes in the coverage of holdings for which records are sent to Eurostat
No changes.

 

3. Changes of definitions and/or reference time and/or measurements of characteristics
No changes.

 

4. Changes over time in the results as compared to previous FSS, which may be attributed to sampling variability
By combining both sample data and administrative data, a Census has been created for some of the characteristics in FSS2016.

 

5. Common land
5.1 Possible changes in the decision or in the methodology to collect common land
Utilised Agricultural Area (UAA) data for FSS reference years prior to 2010 has always excluded Common Land. To address this, an estimate of 422 415 hectares of common land was submitted to Eurostat in February 2011, pertaining to FSS 2007. Therefore the total UAA (including Common Land) for 2007 is now estimated at 4 558 941 hectares. The same total of 422 415 hectares was transmitted in respect of FSS 2010 and a total of 423 020 hectares has been transmitted in respect of FSS 2013. Now in 2016 a total of 427 848 hectares was transmitted as common land for FSS 2016. Common Land data for FSS 2016, FSS 2013 and FSS 2010 are based on the same methodology.
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
The number of specially-created common land units remains at 27 as provided in FSS2013. The total area has increased from 423 020 hectares in 2013 to 427 484 hectares 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 137 560  139 595  - 1.46%   
Utilised agricultural area (ha) 4 883 643  4 959 453  - 1.53%   
Arable land (ha) 458 291  1 041 966  - 56.02% 

Big difference in arable land as temporary grass has reduced significantly from 2013. Temporary grass methodology has changed.

Cereals (ha) 280 370  307 841 - 8.92%  Data completely from administrative sources, which have shown a decrease over the period.
Industrial plants (ha) 11 090   16 421 - 32.46%  Industrial crops dropped to 2010 level. Also industrial crops are now determined completely from administrative data. 
Plants harvested green (ha) 125 691  670 813  -  81.26% Improved methodology to determine temporary grass. Break in series between FSS 2013 and FSS 2016 for temporary grass. 
Fallow land (ha) 3 291  14 295 - 76.98%  Improved methodology to determine fallow land. Break in series between FSS 2013 and FSS 2016 for fallow land. 
Permanent grassland (ha) 4 423 591  3 915 773  + 12.97%  Large increase in permanent grassland, some temporary grass in 2013 is permanent in 2016.
Permanent crops (ha) 1 686  1 594 + 5.77%   
Livestock units (LSU) 6 195 639  5 929 362  + 4.49%   
Cattle (heads) 7 222 116  6 902 646  + 4.63%   
Sheep (heads) 5 140 413  4 942 222  + 4.01%   
Goats (heads) 9 242  10 199  - 9.38%   
Pigs (heads) 1 603 899  1 551 715  + 3.36%   
Poultry (heads) 11 053 009  10 133 159   + 9.08%   
Family labour force (persons) 246 649  252 271 - 2.23%   
Family labour force (AWU) 147 498  150 476  - 1.98%   
Non family labour force regularly employed (persons) 18 528  17 236  + 7.5%   
Non family labour force regularly employed (AWU) 9 999  10 133  - 1.32%   
8.2.1. Length of comparable time series

[Not requested]

8.3. Coherence - cross domain
1. Coherence at micro level with other data collections

At micro level data was examined throughout the editing process. Different data sources were used to evaluate the data that was to be processed. Sources included the FSS 2013, Census 2010 and administrative sources from the Ministry. 

 

2. Coherence at macro level with other data collections

Wherever possible, FSS data were also compared with other available sources and data in other domains. For example the results were compared with FSS 2013, Census of Agriculture 2010 as well as annual crop and animal production surveys. 

Examination of the animal production data showed some very slight differences which were deemed acceptable. Production of crops data show some differences with the FSS 2016 results which can be explained by differences in definitions and/or reference periods and as such can be of limited use. 

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

not applicable

9.1. Dissemination format - News release

[Not requested]

9.2. Dissemination format - Publications
1. The nature of publications

1. Some results for June 2016  were published nationally on September 8th 2016 as "Crops and Livestock Survey June 2016 Provisional estimates". This publication covered cereals, potatoes, cattle and sheep at national level.

2. Final areas of crops and numbers of livestock for June 2016 were published on March 10th 2017 as "Crops and Livestock June 2016 Final Results". This publication provided details of crops at national and regional level and cattle at national, regional and county level, and results for sheep, pigs and other livestock at national and regional level. Background notes are provided with the data.

3. A final FSS publication detailing farms by size, type, economic size and detailed farm labour force data will be published once NMR and FSS2016 dataset have been validated and accepted by Eurostat . The data tables will be accompanied by background notes on data collection, derivation of farm typology, livestock unit coefficients and a copy of the questionnaire.

 

2. Date of issuing (actual or planned)

1. Published nationally on September 8th 2016;

2. Published nationally on March 10th 2017;

3. To be published mid-2018.

 

3. References for on-line publications

1. http://www.cso.ie/en/releasesandpublications/er/clsjp/cropsandlivestocksurveyjuneprovisional2016/

2. http://www.cso.ie/en/releasesandpublications/er/clsjf/cropsandlivestocksurveyjunefinal2016/

9.3. Dissemination format - online database
Dissemination format - online database

 1. AAA08 Number of cattle in June by type of cattle, region, county and year:

http://www.cso.ie/px/pxeirestat/Statire/SelectVarVal/Define.asp?maintable=AAA08&PLanguage=0

2. AAA07 Number of livestock in June by type of animal, region and year:

http://www.cso.ie/px/pxeirestat/Statire/SelectVarVal/Define.asp?maintable=AAA07&PLanguage=0

3. Area Farmed in June by type of land use, region and year:

http://www.cso.ie/px/pxeirestat/Statire/SelectVarVal/Define.asp?maintable=AQA05&PLanguage=0

9.3.1. Data tables - consultations
Data tables - consultations
 Not available. 
9.4. Dissemination format - microdata access
Dissemination format - microdata access

There are usually only a very small number of requests for access to microdata, if any. Applicants must submit a detailed application through the Research Microdata File (RMF) application which requests comprehensive details around the researcher, the purpose of the research and the proposed outputs. The application goes through several layers of approval with the final approval always resting with the Director General of the CSO. Where approved, data may not be taken off-site and is accessed on a stand-alone PC with no network or internet connections. All outputs are placed in a folder to be examined by the Data Custodian to ensure there are no breaches of confidentiality.

9.5. Dissemination format - other

[Not requested]

9.6. Documentation on methodology
1. Available documentation on methodology
National Methodological Report (as requested by Eurostat) is the main methodological document for FSS.

 

2. Main scientific references
Not available.
9.7. Quality management - documentation
Quality management - documentation
 National Methodological Report as requested by Eurostat.
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

While the FSS data is not collected in another CSO survey, the annual June agriculture survey is replaced by the FSS in FSS years.

The utilisation of Bovine Register and IACS register eliminates the need for farmers to provide this data.

The CSO is focused on continuously reducing the response burden on farmers. The final section of the FSS questionnaire, and indeed every agricultural survey, asks the respondent to indicate, in minutes, how long it took to complete the form. This allows CSO to measure the change in response burden from year to year.


11. Confidentiality Top

not applicable

11.1. Confidentiality - policy
Confidentiality - policy

All information returned on FSS questionnaires is treated as strictly confidential and is used for statistical purposes only. This is guaranteed by both Irish and EU law.

Section 33 of the Statistics Act 1993 states:

33.(1) No information obtained in any way under this Act or the repealed enactments which can be related to an identifiable person or undertaking shall, except with the written consent of that person or undertaking or the personal representative or next-of-kin of a deceased person, be disseminated, shown or communicated to any person or body except as follows:

( a ) for the purposes of a prosecution for an offence under this Act;

( b ) to officers of statistics in the course of their duties under this Act;

( c ) for the purposes of recording such information solely for the use of the Office in such form and manner as is provided for by a contract in writing made by the Director General which protects its confidentiality to his satisfaction. 

The Act guarantees the confidentiality of all data provided, expressly prohibiting the disclosure of information which can be related to any identifiable person or enterprise. (See Part V - Protection of Information of the Act above). It specifies the offences and penalties occurred for breaching this confidentiality. (See Part VI - Offences, Penalties and Evidence of the Act above).

11.2. Confidentiality - data treatment
Confidentiality - data treatment

In the national release, a category is primary confidential if any one of the following conditions applies:

  • there are three or less units,
  • one unit accounts for more than 80% of the total (dominance rule 1),
  • two units account for more than 90% of the total (dominance rule 2).

 A category is secondary confidential if publishing that category indirectly reveals information about a confidential category.


12. Comment Top
1. Possible improvements in the future
FSS and Census questionnaire will be re-designed for 2020 to take account of new data needs in IFS Regulation.

 

2. Other annexes
Not available.


Related metadata Top


Annexes Top