This document discusses Italy's strategy for transitioning from a traditional decennial census to a continuous census model that integrates statistical and administrative data. It describes Italy's 2011 census which first began combining census data and administrative records. Experimental surveys in 2015 further tested registering the population based on administrative records supplemented by sample surveys. The goal is a continuous census system that uses statistical registers built from integrated administrative data sources and sample surveys to regularly update and estimate population counts and characteristics.

1. The integration of statistical and administrative
data sources to increase population census and
social data availability
Transforming the census – to 2021 and beyond,
University of Winchester, Winchester, UK
Angela Chieppa, Gerardo Gallo, Donatella Zindato
Italian National Institute of Statistics (Istat)
Unit for Population Census and Integration of Social Surveys)

2. The Population Census before 2011
A decennial population census has been taken in Italy
since 1861
 complete field enumeration (so called “door-to-door”
enumeration)
 census forms delivered and collected by enumerators
and self-filled in by respondents
 all information collected and processed on a complete
basis
 same economic, human and organisational resources
were allocated to every household

3. The 2011 census strategy
Combining the study of census experiences of other
countries with a more effective use of administrative data
held by Municipality Population Registers (MPRs) 
completely new strategy
methodological and technical innovations
adoption of standardised solutions according to
municipality size (necessity to differentiate census
organization according to the needs and capacities of the
different actors)
use of new territorial instruments designed to improve
coverage and quality of the enumeration

4. From a “door-to-door” to a register-supported census
 completeness and simultaneity of the population count
 fieldwork guided by population registers (the 8,092
MPRs have been used as lists of households, and
addresses, to which census questionnaires have been
mailed out)
 self-completed questionnaires have been collected by a
multi-mode system which included
 Internet
 return at any post office in Italy
 return to Municipal Collection Centres
 targeted recovery of non-response by enumerators
 short form/long form strategy in biggest municipalities
(those with at least 20.000 inhabitants + all province seats)

5. Lessons learnt
 Spontaneous return went very well
 Web return (34,5%) was much higher than expected (no
particular emphasis in the communication campaign)
 Logistics of the mail out/mail back process proved very
complex and with many potential (and actual) points of failure
 Not all addresses in the municipal registers could be processed by the
contractor system
 Many questionnaires could not be delivered for various reasons (about
2,000,000 had to be delivered directly by the enumerators)
 Information about questionnaires returned to post offices was not always
reported promptly, making it impossible for municipal offices to keep track
of which households had to be contacted to prompt completion of the
questionnaires

6. Still need for a change
In spite of their improvements, the innovations designed
for 2011 census are not enough to achieve a stable and
enduring balance between census costs and benefits.
Costs remain high and too concentrated in time, while
the use of administrative data is not tested to the
potential offered by the Italian context.
Census data continue becoming quickly out-dated, and
the supply of highly detailed geographic data remains
only decennial.
 A totally different approach seems necessary,
combining a most effective use of administrative sources
with the spread over years of the census fieldwork.

7. EXPERIMENTAL SURVEYS 2015
Aim = test of methodologies, enumeration techniques and organization to be used for
the Continuous Census  INTEGRATED USE OF ADMINISTRATIVE SOURCES
AND SAMPLE SURVEYS
 Main administrative source: Municipal Population Registers
2 separate surveys with separate aims
 C-Sample survey = statistical test on the entity of the coverage errors of
Municipal Population Register with the aim of determining legal population
(Hypothesized strategy for the population count  if coverage error is < an
acceptable threshold, Municipal Register population = legal population; if
coverage error is > predetermined threshold, the Municipality would have to
review and update the Population Register - the field-check conducted by the
Municipality in order to update the population register would be guided by
signals derived from integrated administrative sources)
 D-Sample survey = rolling survey to collect information on population socio-
economic variables (data not available in administrative sources); it would
replace the 2011 Population Census long form in order to produce hypercubes
required by EU Regulation and census data traditionally provided to Italian
data users

8. … a bridge towards change to reduce costs
Annual
basis
Integrated
administrative
data
Estimation
models
Improve quality by
way of integrated
social surveys
system

9. Modernization project towards a
Register based statistics… continuously updated
• Production and use of a
statistical population
register
• Base-registers linked at
the micro-data level by
using a unique ID number
• Continuous update from
administrative sources
and integrated social
surveys system
Population Activities
Territory Business

10. 10
Beyond the stovepipe model =
NSIs traditional organization of data production processes
where each survey (including the Population Census) is
conducted independently from the others
This may cause:
• inconsistencies between data coming from different
surveys due to differences in enumeration/edit and
imputation/estimation processes
• production of less accurate estimates (not all available
information - including information from administrative
sources - is used)
Need to redesign social surveys, also with a view to the
Continuous Census =
The different surveys on the same population of interest (e.g.
individuals and households) are considered just parts of the
same statistical system

11. 11
Modular System:
 General Module: demographic variables for which
information is collected in every survey
 Harmonized modules: variables for which
information is collected in more than one survey
 Specific modules: for which information is collected
in just one survey
Project on the integration of
European Social Surveys

12. To create a Master Sample Data Survey to assess basic
population register (of individuals and households) and
produce population census outcomes
Exploitation of administrative data to create statistical
registers in support of population census
To improve quality of outcomes in terms of timeliness and
completeness … enabling the stakeholders more updated
outputs on different demographic events and for specific
subpopulations
Integration towards the Continuous Census

13. Complete rethinking of the
Continuous Census strategy
 C-aim = produce population counts by
integrating sample survey data and statistical
registers data
 D-aim = production of hypercubes for both EU
and national stakeholders
 Assessment of available information from administrative
sources with a view to replace field collection
 Integration of current social surveys - possible use of ad
hoc module within the Master Sample Survey to collect
census information not available from other sources
(statistical registers and current surveys)


14.  Population Register has been integrated with other
administrative sources (by using a unique ID number):
 Labour Registers
 Education Registers
 Retired and Non-Pension Benefits Registers
 Tax Returns Register with indirect signals (for people declared
as dependents = spouses or children)
 Permits to Stay Register
Each record in administrative data is treated like a signal of presence in
Italy under the definition of usual residence EU Regulation
(1260/2013) which has been evaluated in a period of two years:
January 2012-December 2013 (31st December 2012 represents the
reference date)
C-AIM (a trial within the C-aim)
Activities with direct signals

15. Time period from January 2012 to December 2013 Profile of presence in employment and educational registers
Ja F M A M Jn Jl A S O N D Ja F M A M Jn J A S O N D
1 Steady
2 Outgoing signals with presence in December 2012
3 Ingoing signals with presence in December 2012
4 Signals of presence around December 2012
5 Not steady, at least 12 months
6 Seasonal signals
7 Less than 12 months, not seasonal
8 Just one signal of presence in December 2012
9 Presence only before December 2012
10 Random signals only before December 2012
11 Ingoing signals after December 2012 and presence till 2013
12 Ingoing signals after December 2012 but not till December 2013
Monthly presence scheme of continuity’s patterns in job and
study activities (direct signals)

16. Step I Step II Step III
PR LE PS
Group Count
RR-NPR
Group Count
TR
Group Count61,068 37,704 3,378 20,764 26,649
Signals Signals Signals
Yes Yes - A 36,618
Yes No - B 24,450
Yes E 14,485
No F 9,965
Yes G 6,939
No H 3,026
No Yes - C 1,086
No No Yes D 351
Table legend:
Eligible as usually residents
Uncertain residents
Administrative sources legend:
PR Population Register
LE Labor and Education Registers
PS Permits to stay Register
RR-NPR Retired and Non-Pension Benefits Registers
TR Tax Returns Register
Proceeding scheme and counts of population groups (000)
People in PR with NO signals in other sources
People with signals only in administrative
sources
People with only Permits to stay
People in PR with signals in other sources
Retired people
Dependents
About 410,000 People with strong –
steady signals of usual residence

17. Italian citizens (about 2,2 million people)
(Values per 1,000 Italian people)
Foreign citizens (about 800,000 people)
(Values per 1,000 Foreign people)
Subpopulation H: people recorded in the population register
without signals from other sources

18. Italian citizens
(Values per 1,000 Italian people)
Foreign citizens
(Values per 1,000 Foreign people)
Subpopulation C1: people not recorded in the population
register with strong-steady signals of usual residence

19. Subpopulation C1: Foreign workers with strong-steady
signals and their association between the categories

20. Signals of presence on the territory are extracted from the
sources and can be used effectively to improve the quality of
population registers
ISTAT defined a preliminary workflow to integrate the use of
administrative sources and the official population registers in
order to calculate the usually-resident population
Using this workflow, it is possible to define a group of
individuals eligible to be included in the usually-resident
population of Italy at a given reference date
Conclusions(1)

21. Demographic variables, especially those of gender, age and country
of citizenship as well as the location of the signal on the territory
have proved to be very significant variables for defining specific sub-
population profiles
The division into sub-groups and clusters allows the identification of
subpopulations (for example, the typical foreign communities that
elude the registers, but perform specific labor activities)
These same groups can represent the basis for defining a census
strategy formulated on the use of mixed techniques that combine
specific surveys and appropriate statistical models
The analysis of signal strength based on its continuity over time is
only the starting point in the use of longitudinal data that can be
processed with administrative sources
Conclusions(2)