Diapositiva 1

1
The setting up of Farm Register the production
process to building up the prototype of the
pre-census list
Session 5 Projects on the Improvement of
Business Registers

• By Caterina Viviano (ISTAT- Italy)
• 21 Meeting of the Wiesbaden Group on BR-
• Roundtable on Business Survey Frames

Paris, November 24-27, 2008
2
Contents
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

• Background
• The input sources description
• The production process to build up the prototype
  list pre-treatment and integration of sources
• The pre-census prototype-list the quality
  survey
• Timetable and future work

Paris, November, 2008
3
Background
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

• To enlarge the statistical register system to
  the sector of agriculture (European regulations)
• To exploit all information available into
  administrative sources
• To improve the quality of agricultural statistics
• To introduce innovations concerning aspects of
  the Census survey the use of lists as a support
  for the survey

Paris, November, 2008
4
The project to build up the Farm Register phases
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Paris, November, 2008
5
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
The input sources description

• Specific sectoral sources
• The Integrated Administration and Control System
  (AGEA)
• The System for the Identification and
  Registration of Bovine Animals and other species
  (AA.ZZ.)
• Lands property Incomes (Tax Agency)
• The Land registry
• Agricultural self-employed workers (Social
  Security)
• Regional agricultural Systems.
• General sources
• 7. Chambers of Commerce (CCIAA)
• 8. Fiscal register (VAT)
• Additional statistical sources
• 9. BR
• 10. The Census of Agriculture 2000

Paris, November, 2008
6
The Integrated Administration and Control System
(AGEA)
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
System of files managed by AGEA i.e. the Italian
Paying Agency Recorded unit agricultural
holding obliged to present for any aid
application That unit file is constituted the
first time a request of premium is
presented Updating is made on the basis of
yearly requests of aid

Paris, November, 2008
7
The Integrated Administration and Control System
(AGEA)
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Information content 1 - identification data
Unique Code of Agricultural Holdings (CUAA code)
i.e. the fiscal code of the holder, holders
name, permanent address or place of residence,
VAT number if present, dates of inscription and
updating 2 - territorial data
agricultural parcels managed by the farmer, use
of each parcel (crops, livestock), hectares
invested by type of product, fiscal code of
landholders if different from the holder For the
2006/2007 agricultural season Ident. info 1,9
millions record Territorial info 19.3 millions
record

Paris, November, 2008
8
Lands property income
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Data managed by the Tax Agency Recorded unit
tax payer declaring income derived from
landlords rights and/or agricultural activities

Declaration models and their sections
Paris, November, 2008
9
Lands property income
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

In figures
Type of model Records
   
Natural persons 8,933,520
Simple Partnerships 54,462
Non profit institutions 13,562
Income earned by employees 2,986,031
Paris, November, 2008
10
The System for the Identification and
Registration of Bovine Animals and other species
(AA.ZZ.)
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Data managed by the Ministry of Welfare Recorded
unit animals and their holders with the scope
to preserve public health. The covered animal
species are bovines, pigs, sheep and goats,
poultry, equines

The national database of Bovine Animals realized
in accordance with the EC n. 1760/2000 for the
setting up of a System for the Identification and
Registration of Bovine Animals, is the only fully
operative register up to date
For sheep and goats the existing regulation (EC
n. 21/2004) to set up a register is not yet fully
operative No regulations for others species
Paris, November, 2008
11
The System for the Identification and
Registration of Bovine Animals and other species
(AA.ZZ.)
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
In figures

1) Bovines 239,304 2) Speep and goats 148,894
3) Pigs 89,332 4) Poultry 18,062 5) Equides
29,706
Paris, November, 2008
12
The Land registry
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Unit information 1) parcels referring to land
parcel identification codes, parcel
characteristics like soil quality (i.e. land
devoted to vineyard or sowable ground), hectare,
estimated agricultural income 2) their owners
natural person -name, surname, sex, birth date,
fiscal code juridical person name, fiscal
code 3) their ownerships title lands rights
(property, usufruct), property share and property
settlement.

Paris, November, 2008
13
The Land registry
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

• Many advantages.but also many problems such as
• files organized by municipality
• enormous variety of information difficult to
  treat
• the parcels owner cannot be the holder
• updating is not homogenous
• soil quality could not be updated
• difficult to identify agricultural uses only

Paris, November, 2008
14
summary
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Characteristics of sources units and coverage

Statistical unit compliance
high
high
low
high
high
low
low
Coverage Coverage
theor. real
T P undercoverage
Ts Ts
T T overcoverage
P P
P P
Ts Ts
Ts Ts overcoverage
Source Administrative unit
AGEA Farmer or agricultural holding
AAZZ Livestock holder
Lands income Natural/juridical person
Ch.Com. Enterprise /local unit
VAT Legal unit
Social Security Farmer or agricultural worker
Land registry Parcels owner
TTotal TsTotal for the reference population
Ppartial
Paris, November, 2008
15
The production process
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Phase 1 The pre-treatment of each
administrative file Phase 2 The physical
integration the intra-archive link the link
among different archives Phase 3 The units
identification process
Prototype list

Paris, November, 2008
16
The production process
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

• The integration strategy
• AGEA, used as pivot
• Three initial basic sources AGEA-AA.ZZ.-Lands
  property income, treated according to a linkage
  process that uses all available matching keys so
  that links among records can be classified
• Direct link a match between the primary key of
  AGEA and the primary key of the other source.
• Indirect link a match between the primary key
  of AGEA and the secondary keys of the other
  source.
• Complex link a match between the secondary key
  of AGEA and the primary/secondary key of the
  other source.
• c) Hierarchical integration of the other 6
  sources
• d) Land registry integrated by parcel and owners
  fiscal code

Paris, November, 2008
17
The pre-treatment process
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
AGEA Unit identification Primary key CUAA
Secondary key landowners fiscal code

Paris, November, 2008
18
The pre-treatment process
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Lands property income Unit identification Pri
mary key taxpayers fiscal code Secondary keys
dependent relative and participated firms
fiscal code

Paris, November, 2008
19
The integration process
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Structure of matched records

A potential agricultural unit corresponds to any
integrated record (single record i.e. a
CUAA/Fiscal code) or to any group of records (a
cluster). In the latter case the link among
fiscal codes must be verified on the field.
Paris, November, 2008
20
The units characters
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

• Units attributes identification (i.e location)
  and stratification variables
• taken from input sources i.e. agricultural
  utilized area (hectares)
• derived from deterministic methods indicator of
  existence, eligibility degree, belonging to a
  cluster (familiar links)
• Problems measurement errors, missing data,
  coverage
• Decision No threshold applied, any unit belongs
  to the prototype list BUT with different quality
  i.e. with different signals to be an agricultural
  holding

Paris, November, 2008
21
The quality survey
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

• To test on the field the prototype-list with the
  aim to (re)-define the correct statistical model
  to set up the final pre-census list for 2010
• Checks to evaluate
• the correct use of administrative sources
• the integration methodology
• the units identification
• the criteria for data imputation
• The statistical methodologies to identify the
  census universe

Paris, November, 2008
22
The quality survey
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames

• The prototype list is composed by 7 sub-lists
• the integrated AGEA-based list (1)
• the integrated not AGEA-based list
• Residuals of lands property income with
  agricultural signals (2)
• Residual of CCIAA / VAT with agricultural
  activity, secondary or primary (3)
• Residual of CCIAA without agricultural activity
  but with lands from the land registry (4)
• Residual of CCIAA / VAT with agricultural
  activity but with lands from the land registry
  (5)
• Other lands property income records without
  agricultural signals but with lands from the land
  registry (6)
• Residual of lands from the land registry only (7)
• The probability a unit be an agricultural holding
  decreases starting from list 1.

Paris, November, 2008
23
The quality survey
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Constrains a) 80 municipalities chosen on the
basis of the national agricultural structure b)
120,000 units created by the prototype list c)
Sample size 16,000 Sample design two stages
stratified sample 1 stage chosen
municipalities 2 stage units are chosen
according to an allocation method proportional to
each sub-list size Purpose to select a high
percentage of less elegible (uncertain) units
(70 of the sample) survey technique -
face-to-face interviews carried out in the units
residence place asking information about all
lands and livestock

Paris, November, 2008
24
Timetable and future work
21th Meeting of Wiesbaden Group on BR
Roundtable on Business Survey Frames
Timing First prototype list realization - end
September 2008 (agricultural data referred to
years 2006/2007) Quality survey launched
beginning November 2008 Results of the survey
foreseen - end March 2009 Updated second
prototype list end April 2009 Ex post
reconciliation between survey results and the
second prototype list by June 2009 Pre Census
final list by April 2010 Census of Agriculture
October 2010 AfterThe Farm Register

Paris, November, 2008