COMMISSION STAFF WORKING DOCUMENT IMPACT ASSESSMENT REPORT Accompanying the document Proposal for a REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on European statistics on population and housing, amending Regulation (EC) No 862/2007 and repealing Regulations (EC) No 763/2008 and (EU) No 1260/2013

EN EN
EUROPEAN
COMMISSION
Brussels, 20.1.2023
SWD(2023) 11 final
COMMISSION STAFF WORKING DOCUMENT
IMPACT ASSESSMENT REPORT
Accompanying the document
Proposal for a
REGULATION OF THE EUROPEAN PARLIAMENT AND OF THE COUNCIL on
European statistics on population and housing, amending Regulation (EC) No 862/2007
and repealing Regulations (EC) No 763/2008 and (EU) No 1260/2013
{COM(2023) 31 final} - {SEC(2023) 38 final} - {SWD(2023) 12 final} -
{SWD(2023) 13 final} - {SWD(2023) 14 final} - {SWD(2023) 15 final}
Offentligt
KOM (2023) 0031 - SWD-dokument
Europaudvalget 2023
2
Table of contents
1. INTRODUCTION: POLITICAL AND LEGAL CONTEXT............................................................... 5
2. PROBLEM DEFINITION .................................................................................................................. 11
3. WHY SHOULD THE EU ACT? ........................................................................................................ 24
4. OBJECTIVES: WHAT IS TO BE ACHIEVED? ............................................................................... 26
5. WHAT ARE THE AVAILABLE POLICY OPTIONS? .................................................................... 29
6. WHAT ARE THE IMPACTS OF THE POLICY OPTIONS? ........................................................... 37
7. HOW DO THE OPTIONS COMPARE?............................................................................................ 50
8. PREFERRED OPTION ...................................................................................................................... 53
9. HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?.................................. 58
ANNEX 1: PROCEDURAL INFORMATION............................................................................................ 64
ANNEX 2: WHO IS AFFECTED AND HOW? .......................................................................................... 71
ANNEX 3: ANALYTICAL METHODS ................................................................................................... 124
3
Glossary
Term or acronym Meaning or definition
B2G Business-to-government
BRG European Commission’s Better Regulation guidelines
Census Decennial population and housing census data collection
Census Hub Web tool for central access to European census outputs
Census Regulation Regulation (EC) No 763/2008 on population and housing
censuses
CES Conference of European Statisticians
CFR Charter of Fundamental Rights of the European Union
DEGURBA Degree of urbanisation
Demography Regulation Regulation (EU) No 1260/2013 on European demographic
statistics
DSS Directors of Social Statistics (Eurostat expert group)
EEA European Economic Area
EFTA European Free Trade Association
EP European Parliament
ESA European System of Accounts
ESOP European statistics on population
ESS European Statistical System
ESSC European Statistical System Committee
EU European Union
EUDPR Regulation (EU) 2018/1725 on the protection of natural persons
with regard to the processing of personal data by the Union
institutions, bodies, offices and agencies and on the free
movement of such data (EU Data Protection Regulation)
Eurobase Public database of European statistics disseminated by Eurostat
Eurostat Statistical office of the European Union, Directorate-General of
the European Commission
EU-LFS European Union labour force survey
EU-SILC European Union statistics on income and living conditions
4
FTE Full-time equivalent
FUA Functional urban area
GDP Gross domestic product
GDPR Regulation (EU) 2016/679 on the protection of natural persons
with regard to the processing of personal data and on the free
movement of such data (General Data Protection Regulation)
ISG Interservice group of the European Commission
LAU Local administrative unit
LGBTIQ+ Lesbian, gay, bisexual, transgender, intersex, queer and other
minority gender identities and sexual orientations
MCDA Multi-criteria decision analysis
Migration Regulation Regulation (EC) No 862/2007 on Community statistics on
migration and international protection
MS Member State(s) of the European Union
NSI National statistical institute
NUTS Classification of territorial units for statistics
OI-OO ‘One in, one out’ approach to minimise new net burden for
citizens and businesses
OPC Open public consultation
PHD Private holders of data
REFIT Regulatory fitness and performance programme of the European
Commission
TEU Treaty on European Union
TFEU Treaty on the Functioning of the European Union
UN United Nations
UNECE United Nations Economic Commission for Europe
UNSD United Nations Statistics Division
Usual residence Statistical concept to establish the population base, i.e. the set of
persons (the ‘usually resident population’) that should be
included in population statistics for a given geographic area
5
1. 1. INTRODUCTION: POLITICAL AND LEGAL CONTEXT
Eurostat and the European Statistical System
Eurostat, the statistical office of the European Union (EU), ensures the production of high-quality,
comparable European statistics1
under Regulation (EC) No 223/20092
(the ‘Statistical law’) and the
statistical principles, notably those laid down in the European statistics Code of Practice3
. These
rules and principles aim to ensure, among other things, the independence, impartiality, objectivity
and reliability of European statistics, and through those objectives public trust in the statistics. The
main uses of European statistics are to serve EU policy design and implementation and monitoring,
and their main users are EU institutions.
The European Statistical System (ESS) is the partnership between Eurostat and the national
statistical institutes (NSIs), as well as other national authorities responsible for developing,
producing and publishing European statistics in each Member State. This partnership also includes
the European Economic Area (EEA) and European Free Trade Association (EFTA) countries, i.e.
Iceland, Liechtenstein, Norway and Switzerland. The ESS functions as a network in which
Eurostat’s role is to lead the way in the harmonisation of statistics in close cooperation with
national statistical authorities, which collect data and compile statistics for national and EU
purposes. To fulfil this role, Eurostat issues statistical regulations and methodological guides,
organises expert groups, and assesses the quality of statistics and Member States’ legal compliance.
In line with the EU’s principles of subsidiarity and proportionality, each ESS member develops a
statistical system suitable to their individual institutional context, while still following the common
rules. The ESS also coordinates its work with candidate countries, other Commission departments
at EU level, and international organisations, such as the United Nations Economic Commission for
Europe (UNECE) and other United Nations (UN) bodies.
Eurostat’s activities are further influenced by overarching policies such as the EU’s Better
Regulation agenda4
, which promotes open and transparent EU decision-making and evidence-based
decisions, and European Commission President Ursula von der Leyen’s six political priorities,
among them ‘An economy that works for people’, ‘Promoting our European way of life’ and ‘A
new push for European democracy’5
. Implementing, monitoring and assessing these priorities
requires impartial and objective data – that is, official statistics.
Policies for the EU’s population and citizens
Under Article 9 of the Treaty on European Union (TEU), every national of a Member State, in
addition to the national citizenship, is also a citizen of the EU. To define and implement policies
and activities benefiting the EU population and citizens in the areas of the EU’s powers, as laid
1
Statistics in the ESS context are defined according to Article 3(1) of Regulation (EC) No 223/2009 (see footnote 2) as
‘quantitative and qualitative, aggregated and representative information characterising a collective phenomenon in a
considered population.’
2
Regulation (EC) No 223/2009 of the European Parliament and of the Council of 11 March 2009 on European statistics
and repealing Regulation (EC, Euratom) No 1101/2008 of the European Parliament and of the Council on the
transmission of data subject to statistical confidentiality to the Statistical Office of the European Communities, Council
Regulation (EC) No 322/97 on Community Statistics, and Council Decision 89/382/EEC, Euratom establishing a
Committee on the Statistical Programmes of the European Communities (OJ L 87, 31.3.2009, p. 164).
3
European statistics Code of Practice, revised version endorsed by the European Statistical System Committee on
16 November 2017 (KS-02-18-142).
4
https://ec.europa.eu/info/law/law-making-process/planning-and-proposing-law/better-regulation-why-and-how_en
5
https://ec.europa.eu/info/strategy/priorities-2019-2024_en
6
down in Articles 2 and 3 of the Treaty on the Functioning of the European Union (TFEU), EU
institutions need timely, reliable, detailed, harmonised and comparable European statistics. To
observe the principle of non-discrimination in all its activities and the individual citizens’ rights as
enshrined in Articles 10 and 19 of the TFEU and the Charter of Fundamental Rights of the
European Union (CFR), EU institutions need a reliable and comparable count of the whole
population of the EU. Moreover, the Commission is required to monitor and report on the EU
demographic situation in line with Article 159 of the TFEU. EU institutions must also have
accurate and comparable population figures at their disposal for administrative and procedural
purposes, e.g. qualified majority voting in the Council.
The Treaties oblige the European Parliament and the Council to adopt measures for producing
official statistics where necessary for EU policies (Article 338 of the TFEU). Over the past three
decades, many EU policy areas have seen strongly increasing and evolving needs for complete,
coherent, comparable, reliable and regular European statistics on population, demography and
international migration6
to support evidence-based policymaking. For instance, long-standing
policy needs that already led to the current legislation concerned economic, social and territorial
cohesion, migration (including labour market integration and wider developments) or challenges
related to ageing. Policy needs that have emerged more recently concern more detailed geographic
patterns, including on migration, EU and regional mobility, urban/rural integration at local levels,
monitoring the European Green Deal, and equality and non-discrimination.
European statistics on population
Population and demographic statistics are among the most popular data Eurostat produces, and they
are important for almost every area of policy. For instance, EU economic, social and territorial
cohesion policies, labour market and integration policies, equality policies and the EU’s long-term
economic and budgetary projections have been relying on these statistics over the past two decades.
However, these statistics also provide essential input for public research and informing the society
at large. Finally, population statistics are very relevant to the general public because they describe
facts and events that concern each individual.
6
‘Migration’ refers to the general demographic concept of people moving to live in a different country. It excludes
specific administrative actions related to the entry and stay of non-EU nationals, such as asylum applications, granting
refugee status, border controls and issuing residence permits. Statistics on these areas are specified under Articles 4 to 7
of Regulation (EC) No 862/2007 and have been recently updated.
Figure 1 – Scope of European statistics on population
7
As part of this initiative and as illustrated in Figure 1, European statistics on population (ESOP)
cover official European statistics on population, demographic events and migration, statistics from
the population and housing censuses, and the various indicators based on these statistics. Eurostat
has published statistics in these areas since 1960, when the first survey on the size and structure of
the active population in Member States at that time was introduced. Since then, population statistics
have been produced mainly by taking results from a direct enumeration of the population during
population censuses and interpolating intermediate periods with information on population changes
from civil registration administrative systems (on births, deaths and migration). This means the
burden of production has always been mostly on statistical authorities, except for the censuses that
interviewed nearly the entire resident population, typically at decennial intervals. The ongoing
move from traditional field censuses to combined or even fully register-based modes minimises the
burden of production for the general public and moves it onto the statistical authorities.
Until 2007, Member States transmitted all related data on a voluntary basis, which resulted in
inconsistencies and lack of completeness or timeliness (see evaluation of the current legislation)7
.
Therefore, after a previous intervention these statistics are now based on legislation adopted
between 2007 and 2013. First, Regulation (EC) No 862/20078
(Migration Regulation) set out
requirements on migration statistics in line with the action plan for the collection and analysis of
Community statistics in the field of migration9
. It includes statistics on: (i) immigration to and
emigration from Member States, including from the territory of one Member State to that of another
Member State and between a Member State and the territory of a non-EU country; (ii) the
citizenship and country of birth of persons usually resident in the territory of the Member States;
and (iii) statistics on citizenship acquisitions. Regulation (EC) No 763/200810
(Census Regulation)
then set out common rules for providing comprehensive census data on population and housing in
the EU every 10 years. This ensured compilation of detailed data on specific demographic, social
and economic characteristics of persons, families and households, as well as on housing at a
national, regional and local level. Finally, Regulation (EU) No 1260/201311
(Demography
Regulation) laid down common rules for European demographic data, including data requirements
on population stocks and vital events such as births and deaths. This Regulation also obliged
Member States to provide the Commission (Eurostat) with harmonised data on the total national
population to be used as weights for the qualified majority voting in the Council.
All three regulations specified definitions, data variables and periodicity of statistics. Data
breakdowns (e.g. demographic or geographical) and detailed tabulations were defined mostly in
implementing acts. Eurostat currently collects data from Member States according to data
requirements specified in these statistical regulations. Where the legislation does not cover the EU’s
needs for statistics, Eurostat seeks the agreement of Member States to initiate voluntary data
collections. For instance, statistics on marriages, divorces, legally induced abortions, infant
mortality and loss of citizenship are currently entirely voluntary. Various breakdowns of live births
and deaths (e.g. marital and activity status, educational attainment) and of international migration
7
SWD(2023) 13.
8
Regulation (EC) No 862/2007 of the European Parliament and of the Council of 11 July 2007 on Community statistics
on migration and international protection and repealing Council Regulation (EEC) No 311/76 on the compilation of
statistics on foreign workers (OJ L 199, 31.7.2007, p. 23).
9
Communication from the Commission to the Council and the European Parliament to present an Action Plan for the
collection and analysis of Community Statistics in the field of migration, COM(2003) 179.
10
Regulation (EC) No 763/2008 of the European Parliament and of the Council of 9 July 2008 on population and
housing censuses (OJ L 218, 13.8.2008, p. 14).
11
Regulation (EU) No 1260/2013 of the European Parliament and of the Council of 20 November 2013 on European
demographic statistics (OJ L 330, 10.12.2013, p. 39).
8
are voluntary as well. To produce regional population projections that are important for the regional
and cohesion policies and the European Semester, Eurostat collects voluntary data on regional
international migration and interregional migration in the EU.
In 2020, to respond to data needs in the context of the COVID-19 pandemic, Eurostat and Member
States set up the voluntary data collection on weekly deaths that also enables the publication of
statistics on excess mortality. The evaluation has shown that regular use of voluntary data
collections to address important user needs is not sustainable because timeliness and completeness
of statistics at EU level cannot be guaranteed as needed. Many Member States compile and publish
more extensive national population statistics than currently collected at EU level. However, there
are differences in national practices and data availability.
European statistics on population – the backbone of social statistics
Population statistics are the backbone of all social statistics, as they provide the most accurate and
up-to-date reference information on the entire population and its basic demographic characteristics.
Such a population framework with very good coverage and location information is indispensable
for more detailed annual population estimates, data collections based on samples, and regional
analysis. Population estimates are also needed to obtain per capita indicators in statistics.
Population statistics provide the input for preparing population projections for EU long-term
economic and budgetary projections specifically and EU economic, social and cohesion policies
more broadly.
In 2014, to comply with emerging statistical needs, the Commission (Eurostat) began the process of
modernising social statistics. This led to adopting a common legal framework for European
statistics on persons and households, based on data at individual level collected from samples of
persons and households12
in 2019. This framework is fundamental for laying a coherent foundation
at European level for data collections on samples. This initiative on European statistics on
population is the second core element of this modernisation process. Early high-level support for
the initiative in the ESS was expressed in the 2017 Budapest Memorandum13
, which endorses
action on a flexible response to changing needs, further harmonisation of concepts and definitions,
and expanding annual data collection, including on migration and geographic detail.
European statistics on population and the global data and governance environment
At a global level, there is action to coordinate and harmonise population and housing censuses
through the World Population and Housing Census Programme, which is developed and updated on
a 10-yearly basis under the coordination of the United Nations Statistics Division (UNSD). The
current 2020 decennial programme encourages all countries to hold at least one census between
2015 and 2024 and take into account international and regional recommendations on data quality,
methods, concepts and definitions. The programme is supported by a set of principles and
recommendations for population and housing censuses14
that provides extensive guidance to
countries on the design and implementation of the census. Regional and topical guidelines on the
census and different methodological and implementation issues are produced, such as the
12
Regulation (EU) 2019/1700 of the European Parliament and of the Council of 10 October 2019 establishing a
common framework for European statistics relating to persons and households, based on data at individual level
collected from samples […] (Text with EEA relevance) (OJ L 261I, 14.10.2019, p. 1).
13
https://ec.europa.eu/eurostat/documents/13019146/13237859/FINAL+Budapest+memorandum.pdf/96a6db89-1395-
44a5-8a46-85e8c49d576c
14
UN (2017) Principles and Recommendations for Population and Housing Censuses, Revision 3.
9
Conference of European Statisticians’ (CES) recommendations for censuses of population and
housing15
. These CES recommendations are compatible with the world census programme but are
focused on the specific circumstances and needs of countries in Europe. Looking ahead to the next
2030 worldwide census round, the foundations of which are currently being prepared, international
recommendations are shifting further towards more efficient and versatile production systems, for
instance with the 2018 UNECE guidelines on the use of registers and administrative data for
population and housing censuses16
and with the UN Handbook on Registers-Based Population and
Housing Censuses17
(draft in progress). These developments also address lessons learned from the
COVID-19 pandemic, which showed that modern official statistics systems are needed that
minimise dependencies on responses from individual contacts while becoming more efficient to
react to ad hoc public needs.
Demography and migration statistics are other areas for which there is fruitful international
statistical cooperation. The UNSD issues guidelines on demography, vital events and migration
statistics18
, and collects data leading to the publication of the annual Demographic Yearbook19
.
Eurostat has integrated the UNSD data request in its data collection, thus helping 43 European
countries to avoid transmitting the same data twice to international organisations. In the area of
migration statistics, Eurostat also has a close partnership with the International Organisation for
Migration, the United Nations High Commissioner for Refugees and UNICEF on methodological,
promotional and technical assistance matters20
.
Furthermore, official statistics are being developed and produced in the context of a public data
environment that is currently evolving rapidly, given the ongoing digitalisation of all sectors of
public life. More and more often over the past two decades, statistical office complement traditional
methodological approaches with opportunities offered by Big Data to increase topical relevance,
timeliness of data and efficiency. This initially started with the increasing use of administrative
register information that is becoming richer and more accurate – a transformation that is still
ongoing and not equally achieved in all Member States. In addition, more and more new sources
are becoming available (e.g. satellite, social network and mobile operator data), including sources
held in the private sector. A modern governance framework including official statistics should
embrace all these developments. For instance, the recently adopted Commission proposal for a Data
Act21
provides enablers for business-to-government (B2G) data sharing under certain conditions,
including for official statistics. This new opportunity would increase the potential benefits from a
modern data environment and incentivise further modernisation of population statistics22
.
Current issues with European statistics on population
As recognised by the evaluation, statistical data on the European population – including
demographic and migration events, information on families, and households and housing
15
https://unece.org/statistics/publications/conference-european-statisticians-recommendations-2020-censuses-
population
16
https://unece.org/guidelines-use-registers-and-administrative-data-population-and-housing-censuses-0
17
https://unstats.un.org/unsd/statcom/53rd-session/documents/BG-3e-Handbook-E.pdf
18
https://unstats.un.org/unsd/demographic-social/standards-and-methods/
19
https://unstats.un.org/unsd/demographic-social/products/dyb/
20
e.g. in the framework of the Expert Group on Refugee and Internally Displaced Persons Statistics and the
International Data Alliance for Children on the Move.
21
COM(2022) 68.
22
C Bosco et al. (2022) Data Innovation in Demography, Migration and Human Mobility, EUR 30907 EN,
Publications Office of the European Union, Luxembourg, ISBN 978-92-76-46702-1, doi:10.2760/958409, JRC127369.
10
arrangements – are vital for evidence-based policymaking. In the European context, high-quality
statistics across Member States are essential to support many policy areas and initiatives across the
EU. Apart from the long-standing use cases mentioned above, four of the six Commission priorities
for 2019-2024 (footnote 5) have expressed clear needs for specific relevant European population
statistics as data evidence for policy, namely ‘A European Green Deal’ (renovated, energy-efficient
buildings), ‘Promoting our European way of life’ (fundamental rights), ‘A new push for European
democracy’ (long-term vision for rural areas, rights of persons with disabilities) and ‘An economy
that works for people’ (‘Union of equality’). Moreover, the final proposals of the Conference on the
Future of Europe23
have noted a need for further EU-level efforts to collect such data.
Appropriate data evidence for EU policies should be complete, coherent and comparable at EU
level as well as timely and frequent enough to support policymaking effectively. A long-standing
practical example is cohesion funding, which is allocated on the basis of regional population size
and demographic patterns. If the underlying statistics are not comparable across all EU regions, this
can distort the allocation of funding. Therefore, it is important that the data collected across
Member States are coordinated and consistent to have relevant and comparable data to support
policy initiatives and comply with the EU’s aims and objectives. This includes ensuring consistent
statistical definitions and data collection methods, including the periodicity, timeliness,
completeness and required detail of statistical publications at EU level.
In terms of periodicity and timeliness as set out in the current legislation, Eurostat’s demographic
and international migration data collections are mostly annual and available in full detail 12 months
after the reference year. This was insufficient in relation to the COVID-19 pandemic, for instance,
when timelier data were needed. Eurostat currently has no data available to answer user questions
about EU citizens returning to their home countries as a result of Brexit or COVID-19 until the
annual data become available. In addition, a very recent ad hoc collection of statistics on Ukrainian
residents in the EU directly before Russia’s unprovoked invasion of Ukraine showed that a majority
of, but by far not all, Member States could respond with the necessary details on time. Furthermore,
EU-level statistics on population and housing censuses are made available only 27 months after the
reference year. This delay is considered to be too long by key policy users, even while they
acknowledge the complexity of the process in some Member States. There are time lags between
the releases of national census results and EU census outputs as well as across Member States.
While measures were taken to align concepts and definitions in the current legal framework, the
population base is defined in each regulation with some statistically important differences in the
formulation. Moreover, most of the legal definitions currently in place offer default exceptions for
Member States to resort to national population definitions where the harmonised concept of usual
residence based on 12 months’ presence is difficult to determine.
The evaluation has shown that this lack of harmonisation reduces the relevance, coherence,
consistency and comparability of statistics across Member States, with negative impacts for
decision-making based on them. There is therefore the need for a new legal basis that can provide a
long-term framework for necessary developments towards further convergence of population
statistics. This should include real progress on harmonisation and sufficient flexibility to adapt
better to evolving policy needs and to make the most of opportunities emerging with new data
sources. Since there are also potential opportunities for administrative simplification and process
23
‘Future of Europe: Conference Plenary agrees final set of proposals’ (press release IP/22/2763); see measure 15.10.
11
integration compared to the status quo under three non-aligned legal bases, this initiative is
included in the 2022 Commission work programme as a REFIT initiative.
2. 2. PROBLEM DEFINITION
2.1. 2.1. What are the problems?
The evaluation identified issues structured around the following four main problems. The current
legal framework:
 does not ensure sufficiently complete, coherent, and comparable statistics, which may lead
to suboptimal decision-making (Problem 1);
 does not ensure sufficient availability of population data in terms of frequency and
deadlines for data transmission (Problem 2);
 fails to capture characteristics and details of politically relevant topics or groups, e.g. the
data provided to users, and the way it is presented, is not detailed enough in terms of
statistical topics, characteristics, and breakdowns for the population in general and relevant
specific groups of interest (Problem 3);
 is not flexible enough to adapt to evolving policy needs and to enable using data from
administrative and other new sources in Member States and at EU level (Problem 4).
The next section explains these problems in turn and how significant they are in terms of scale and
impact. It also examines the main drivers leading to the problems and reflects on how these
problems are likely to evolve in the absence of targeted action.
2.2. 2.2. What are the problem drivers?
2.2.1. 2.2.1. Problem 1: The current legal framework does not ensure sufficiently
complete, coherent, and comparable statistics, which may lead to suboptimal
decision-making
The most significant quality gap identified by the evaluation in the current legal framework is the
lack of harmonisation of the population base. More precisely, three conceptually different
definitions for resident population are currently allowed in varying formulations across the three
base acts: either strict usual residence based on 12 months’ presence, or in default registered
residence, or legal residence where strict usual residence cannot be determined. The default options
were originally introduced in the legislation to accommodate specific constraints of a small number
of Member States. However, as more and more countries are relying on administrative data sources,
such non-harmonised default definitions are becoming an increasingly common practice across
Member States, sometimes even using different definitions for different datasets. Figure 2
illustrates this fragmented landscape across the ESS. Moreover, the legislation is not detailed
enough to set out exhaustively what is included in (and excluded from) the population.
An entailed gradual loss of relevance had already been accepted when the current legislation was
adopted, since many policy and democratic representation considerations, notably at EU level,
require a population base definition (who is counted among the population and who is not) that
reflects the actual population present. The compromises detailed above have generally led to a
12
barely sufficient situation in this respect. The Demography Regulation only requires national
population figures based on strict usual residence for Council voting weights. However, these
figures must be of the highest possible quality24
but the factual accuracy or comparability achieved
is currently not quantified, and the evaluation indicates that the strict definition for this purpose is
not implemented consistently in all Member States. (This touches on another issue of reduced
completeness, namely insufficient metadata being published along with the statistics, which was
also raised by users during consultation surveys and interviews).
The harmonisation gap also entails a situation where the vast body of demographic and migration
statistics cannot reach its potential in terms of comparability between Member States and
consistency between datasets due to differences in the definitions applied. For instance, different
population bases are often used for population stocks and migration flows, which leads to
inconsistencies between stock differences and demographic changes (demographic balance)
between reference years. During the current legislation’s implementation, only between 8 and 14
ESS members have been reporting consistent demographic balances. Regarding comparability, the
differences between total populations reported for Council voting weights (strict usual residence
nominally enforced) and in other annual datasets (no common population base enforced) may be
used as a rough proxy indicator for the remaining level of comparability limitation across Member
States. From 2014 to 2020, relative differences over the years were on average smaller than 1% (for
23 of 27 Member States), and the largest relative differences encountered for single Member States
were below 5% for any reference year25
.
The open public consultation26
(OPC) has revealed a complex picture around this harmonisation
gap and its relation to coherence and comparability. When asked whether the current statistics are
sufficiently harmonised, comparable and/or coherent overall, OPC respondents across all key
24
A qualified majority under Article 16(4) of the TEU requires at least 15 Council votes representing at least 65% of
the EU’s population, where some (rare) combinations of Member States can lead to results that are extremely close to
the population threshold. For instance, Austria, Belgium, Bulgaria, Croatia, Cyprus, Germany, Hungary, Ireland, Italy,
Latvia, Lithuania, Luxembourg, Poland Portugal and Spain (15 MS) would represent 65.000032% of the EU’s
population based on 2021 data.
25
This assessment only addresses national population totals. It is likely that comparability losses are much bigger in
certain parts of the population that are hard to capture and/or where the coverage is very sensitive to the population
base.
26
https://ec.europa.eu/info/law/better-regulation/have-your-say/initiatives/12958-European-statistics-on-population-
ESOP/public-consultation_en
Figure 2 – Population base definitions currently in use across ESS members, according to responses to the NSI survey. (Source:
ICF analysis of NSI survey Q4.2.1)
13
stakeholder groups did not have a strong tendency to either agree or disagree27
. Across all
stakeholder groups, however, there is a lack of knowledge about the current harmonisation gap of
the population base definition. Only among researchers, the majority of respondents is aware, but
for example in the institutional user group, just about half of the respondents know about this gap.
Then, contrary to generic opinions above, once respondents have been made aware of the current
harmonisation gap, more than 50% across all stakeholder groups strongly agree that harmonisation
would be important28
.
The survey of national statistical institutes (NSIs)29
and in-depth interviews highlighted that NSIs
consider their national definitions to be adapted to the national context and that the benefits of the
current use of national definitions centre around meeting current national requirements. Some NSIs
asserted that this can lead to not entirely comparable statistics at EU level, which can be an issue
for some data users with more advanced needs for comparable population statistics (those for whom
precision at a granular level is required). The use of national definitions can also lead to double
counting when people move between Member States, which may result in discrepancies in
European population statistics. Data users and international partners of Eurostat identified similar
problems, for instance when counting people who migrate between Member States to study or
work, or where people have second homes and spend parts of the year living in two or more places.
Figure 3 illustrates NSIs’ (i.e. statistics producers) opinion that the status quo means accepting a
trade-off between certain drawbacks for the quality and use of the data at EU level (considered
mostly minor impacts) and advantages in terms of reduced production costs of these statistics
(considered mostly a major impact).
The lack of rules on time series revision is another gap leading to patchy revision practices across
Member States and thus to reduced comparability over time, in particular for Member States
relying on more traditional data collection approaches (as opposed to administrative source-based
approaches). This is the case for instance when annual data is computed as incremental changes
(vital and migratory events accumulating over time) since the last census year and thus suffer from
27
Responses to OPC survey (Q2.4).
28
Responses to OPC survey (Q2.5).
29
Responses to NSI survey (Q4.2.2).
Figure 3 – NSI survey respondents’ views on the impact of maintaining different national population base definitions. (Source: ICF
analysis of NSI survey Q3.3)
14
accuracy decreasing with the time gap after the last census. However, the current legal framework
does not contain any rules on time series revisions of annual data. The 2011 EU census experience
illustrates some consequences, where, on Eurostat’s initiative, 18 countries voluntarily revised parts
of their annual time series with average correction magnitudes at total population level of around
1.4% but up to 7.5% in individual cases30
. However, two of these countries only revised back to the
reference year of 2011 (Germany; avg. -1.87%) or 2010 (Poland; avg. -0.81%). This meant that
breaks in their time series remained around the census year 2011, with the German case extending
to a visible time series break of the entire EU population due to the population size of the country.
Consequently, most OPC respondents across almost all key stakeholder groups give high priority to
potential future improvements leading to better revision rules31
.
Finally, continuing large parts of the statistical outputs under voluntary collections leads to various
gaps in quality and other related gaps documented in the evaluation. In particular, a key
consequence is the lack of completeness of voluntary statistics across all Member States (Figure 4).
This limits the utility for EU-level analysis or publication of EU aggregates, and thus reduces the
EU added value significantly. A related issue is the reduced efficiency of this practice: the overall
baseline production costs across Member States delivering voluntary statistics fail to deliver
significant benefits at EU level. Even if more Member States start producing a given voluntary
dataset, the practice will not be cost effective as long as there are other Member States that continue
not to produce this dataset. These findings are in line with the opinion of OPC respondents across
all stakeholder groups (except statistics producers) that potential future improvements should
include measures to regulate data that is currently voluntary32
. Current completeness and
comparability gaps are seen as most critical in the datasets on international migration – from/to and
within the EU.
30
The limitation of footnote 25 applies here too. In particular, a 2014 Eurostat analysis showed that the size of revisions
can be much larger in certain population groups, e.g. only in the basic demographic breakdowns by age and sex
corrections up to +21% were found (for males aged 20-24 in Ireland).
31
Responses to OPC survey (Q3.16).
32
Responses to OPC survey (Q3.13, Q3.14).
Figure 4 – EU level completeness of breakdowns over time: those that became mandatory under the current legislation and those
that are still voluntary today. (Source: Eurostat analysis of Eurobase datasets)
0,0%
20,0%
40,0%
60,0%
80,0%
100,0%
Mandatory Voluntary
15
2.2.2. 2.2.2. Problem 2: The current legal framework does not ensure sufficient
availability of population data in terms of frequency and deadlines for data
transmissions
Currently, mandatory demographic and international migration statistics are published with an
annual frequency and a default deadline of 12 months after the reference period (timeliness) for the
full details. These annual data are complemented by detailed decennial EU outputs from population
and housing censuses that are published based on national schedules within a deadline of 27 months
from the end of the census year (e.g. by March 2024 for 2021 EU census outputs)33
.
The stakeholder consultation has revealed that the current frequency and timeliness of the
publication of statistics remains below user expectations. There were no strong opinions from OPC
respondents across all stakeholder groups (and only a minority of respondents among institutional
users found that the current statistics were generally not timely enough). However, population and
housing census outputs were mentioned most often as most critically affected by insufficient
frequency and especially timeliness34
. The targeted consultation of Commission departments
(representing EU-level institutional users) also indicated critical gaps. In particular, regarding
frequency, various EU policy areas (e.g. rural integration, seasonal movements between rural and
urban territories including workers) have been identified as requiring at least quarterly data on
population stocks and demographics, including migratory events. Moreover, monitoring the
European Green Deal requires housing data more frequently than every 10 years, including data on
energy characteristics of housing35
.
Similarly, responding effectively to disruptive events or crises – requiring effective measures for
quick and very frequent (e.g. monthly or weekly) ad hoc data collections – has become more
33
Whereas 15 Member States already publish national population and migration estimates monthly or quarterly, and a
majority of Member States publish national census results much faster than within the 27 months at EU level.
34
Responses to OPC survey (Q2.4).
35
Responses to the targeted consultation of statistical correspondents of Commission services.
Figure 5 – Number of ESS members currently publishing national data with annual frequency (left) and with infra-annual
frequency (right), by statistical topic and by timeliness. Source: Eurostat analysis of information from NSI national websites (state
as of 08/2021).
0
5
10
15
20
25
T+6 m T+9 m
0
5
10
15
20
25
T+30 d T+45 d T+60 d T+75 d T+90 d
16
important during the past decade. For instance, Eurostat has not been able to sufficiently answer
user questions on EU citizens returning to their home countries following Brexit or the onset of the
COVID-19 pandemic until the standard annual data became available, and provided that Member
States sent voluntary breakdowns. An ad hoc data collection on weekly deaths was initiated to
address COVID-19 information needs, but OPC responses indicate that generally such (voluntary –
see Section 2.2.1) ad hoc measures may not be sufficient to meet such highly dynamic crisis
response needs36
. These issues are also linked to the lack of flexibility in the legal framework (see
Section 2.2.4).
Finally, from the statistics producers’ perspective, the frequency with which NSIs publish
population statistics nationally varies significantly37
. However, Figure 5 shows that many of them
already publish some statistics topics at national level more frequently and/or more quickly than
Eurostat does at EU level. For instance, 22 ESS members publish national population stock data
within 6 months of the reference date, which is 6 months earlier than the 12-months legal deadline,
and at least 17 ESS members manage to do so for data on vital events and on migration as well. The
use of administrative data sources in particular makes more frequent/timely publication less costly
and more efficient. In a poll carried out during a consultation workshop with directors of NSIs38
,
64% of respondents indicated improved timeliness as the biggest priority for a future EU initiative.
However, increasing the EU-level frequency and timeliness of data collections would likely have
varying impacts across Member States, including increased administrative burden (and costs) on
some NSIs, with production processes relying less on administrative sources. Nevertheless, there is
a similar overall cost-effectiveness argument at EU level as regards completeness (see
Section 2.2.1) – understanding that the status quo gets less efficient the more Member States use
baseline resources to attain a certain national standard without reaping the EU added value.
36
Responses to OPC survey (Q2.6).
37
Responses to NSI survey (Q2.16).
38
DSS workshop ‘What users want and what modernised population statistics can deliver (inception results)’, virtual
Teams meeting with 50 registered NSI participants, 23 September 2021.
Figure 6 – Views of OPC respondents from all professional user groups (institutional, researchers and other) on the sufficiency of
current data, by demographic or societal change topic and excluding answers ‘don’t know’. Source: Eurostat analysis of OPC
responses to Q2.3
21
20
28
17
15
8
7
3
10
10
5
3
23
20
20
17
10
3
12
10
13
11
8
7
6
7
5
9
15
13
12
11
16
13
13
7
10
5
6
4
8
13
14
10
8
12
12
19
2
1
0
0
1
8
6
1
2
2
6
19
0% 20% 40% 60% 80% 100%
Population size and structure
Life expectancy
Population age
Fertility rates
Household size
Housing and homelessness
EU mobility
Acquisitions and losses of citizenship
Population of urban areas
Population of rural areas
Regional and urban/rural mobility
Groups at risk of inequality or discrimination
Very sufficient
Somewhat sufficient
Neutral
Somewhat insufficient
Very insufficient
17
2.2.3. 2.2.3. Problem 3: The current legal framework fails to capture
characteristics and details of politically relevant groups or topics
The evaluation has identified various gaps in detail that have led to a significantly reduced
relevance of the current legal framework over time from a policy perspective. In addition, a
majority of OPC respondents from professional user groups were of the opinion that, generally
speaking, the statistics do not provide enough detail39
. The most crucial gaps identified revolve
around characteristics of politically relevant topics and groups (migrants, migration and mobility
within the EU; urban/rural population; vulnerable groups for equality, non-discrimination and
fundamental rights policies; energy efficiency of housing for the European Green Deal), as well as
around an insufficient geographic detail of the statistics (especially functional typologies and
georeferenced data for urban/rural mobility and cross-border analysis). These gaps were confirmed
by the targeted consultation with Commission departments, as well as by a majority of OPC
respondents across almost all key stakeholder groups – for instance professional users’ views
shown in Figure 6. Each topic will now be addressed briefly in turn.
Migration and mobility to, from and within the EU
The relevant concept of migration used in ESOP-related statistics is any change of residence at a
given geographic level of detail, where the concept used for ‘residence’ links migration directly to
the population base, or any change thereof, at the same geographic level40
. Therefore, changes of
residence at any geographic level across the EU are in principle considered migration flows within
the scope of ESOP. This includes international migration (between countries) both from/to the EU
and within the EU, as well as internal migration (between regions or other geographic units inside
the same country). Moreover, migrant stocks relate to population subgroups with an immigrant
background (typically identified by a country of birth and/or citizenship different from the country
of residence).
Currently, Eurostat publishes some mandatory data on international migration flows and migrant
stocks under the Migration Regulation (flows and stocks by core demographic characteristics and
broad groups of countries of origin/destination), complemented by a variety of voluntary data
breakdowns of increasing policy relevance, including migration patterns at subnational level, more
detailed and ideally single-country information on origin or destination, and socio-economic
characteristics. However, as the evaluation has shown, most of these voluntary data are affected by
a lack of EU completeness and comparability. Furthermore, there are no European statistics yet for
other specific migratory movements (like short-term/seasonal movements or return migration) that
receive increasing policy attention but are even harder to capture.
Consequently, OPC responses across user-type stakeholder groups indicate an insufficient status
quo; for instance, EU and subnational mobility rank among the most insufficient in professional
users’ views, as seen in Figure 6. Moreover, out of 92 OPC respondents overall that noted any
inadequacies in the current data, the majority identified migration within the EU (66%) directly
followed by migration from/to the EU (62%) as high priorities to improve by 203041
. This is
39
Responses to OPC survey (Q2.4).
40
It is this defining notion linking migration to a change of residence according to a given population base concept that
separates the migration data within the scope of ESOP (Article 3 of the Migration Regulation) from the other statistics
on administrative and judicial events related to international migration, legal and irregular migration under Articles 4 to
7 of the Migration Regulation.
41
Responses to OPC survey (Q2.4).
18
substantiated by replies to the targeted consultation of Commission departments, where various
policy areas signalled the need for more detailed migration data – including flows and stocks by
socio-economic characteristics and by single country of origin/destination (flows) or of
birth/citizenship (stocks) – to monitor the integration of immigrants in labour markets and society.
Migration data also appears to be one of the datasets most requested by national data users, as 79%
of NSIs have indicated migration flows within the EU and from/to non-EU countries among the
most commonly requested topics42
. At the consultation workshop with NSIs (footnote 38),
migration within and from/to the EU was also ranked among the top priorities to be improved by
2030, where migration within the EU in particular poses increased challenges due to freedom of
movement entailing reduced availability of administrative sources. Moreover, the quality of
emigration data may be limited and may underestimate the true scale of the phenomenon as existing
data collection processes may not be sufficiently sensitive to disaggregate emigration by detailed
characteristics.
Specific population groups
The term ‘equality data’ refers to data on population subgroups characterised by attributes that
make them more vulnerable to inequality or discrimination. Articles 10 and 19 of the TFEU list six
grounds for discrimination (variables) that are relevant for EU non-discrimination policies: sex,
racial or ethnic origin, religion or belief, disability, age and sexual orientation. To date, European
equality statistics in the area of population statistics exist for the basic demographic characteristics,
sex43
and age44
. More detailed information is available in other areas of European statistics on
42
Responses to NSI survey (Q2.18).
43
The demographic concept of ‘sex’ in most official statistics standards, including in the ESS, links to the (binary)
biological sex assigned at birth. However, while Articles 10 and 19 of the TFEU literally list ‘sex’ from an equality
perspective, this ground for discrimination is normally understood as also including self-identified gender identity. The
latter is not addressed in any European statistics so far.
44
The demographic concept of ‘age’ is generally already well-covered in current statistics, but sometimes a breakdown
by 5-year age bands is used, which some stakeholders noted does not capture children and youth well (e.g. below/above
Figure 7 – Number of Member States publishing equality data as part of their national 2021 census outputs, by equality
characteristics. Source: Eurostat monitoring survey of NSIs on 2021 census activities (situation as of September 2021, sexual
orientation not asked).
0
2
4
6
8
10
12
14
16
Religious
affiliation
Ethnicity Mother
tongue
Disability Non-binary
gender
Race Other (not
specified)
19
disability, but notable gaps persist on race/ethnicity, religious affiliation, gender identity and sexual
orientation. Therefore, the targeted consultation of Commission departments confirmed that
European equality data are greatly needed across all grounds for discrimination for EU policies on
fundamental rights and non-discrimination, especially in the context of the ‘Union of equality’45
.
Moreover, OPC respondents from professional user groups believe that equality is the topic with
the most insufficient available data (see Figure 6; only 18% of the group that expressed an opinion
finds the current data sufficient). Most recently, the final proposals of the Conference on the Future
of Europe have also noted a need for further EU efforts to collect such data (footnote 23).
At national level in the ESS, the coverage of equality data is patchy as well. For instance, according
to a recent Eurostat survey on implementation of the 2021 census round, only 19 ESS members
published any variables on population subgroups at risk of inequality or discrimination as part of
their national 2021 census outputs46
. Figure 7 shows the number of ESS members publishing
specific equality variables. The most prevalent reasons for not providing certain equality topics
across Member States seem to be historical or cultural perceptions of such data as very sensitive or
not to be collected by public authorities. Moreover, a majority of NSI respondents saw
implementation costs (10 Member States) and national census modes (9 Member States) as a huge
barrier to including self-declared information in their 2031 censuses47
. National legal barriers are
often brought up as an argument as well, but recent Commission studies have found that such
barriers do not factually exist in any Member State on race/ethnicity and gender48
.
National contexts aside, there are technical and feasibility challenges for the collection of such data.
For instance, there is a generally accepted principle that these characteristics should be self-
identified49
, which would require primary data collection (e.g. based on surveys) rather than
extraction from administrative sources (let alone the fact that such characteristics are often not
recorded at all in public administration registers). Moreover, the quality of results from sample
surveys depends on an accurate sampling of the target population. This is another key challenge
because the best sampling frames available are often based (again) on administrative registers50
.
However, these registers have notorious under-coverage issues with certain vulnerable groups –
like Roma or migrants with unclear or irregular status – due to them not being registered in the
national systems.
Housing data for the European Green Deal
Only the censuses currently provide decennial outputs on housing arrangements and homelessness,
including data on housing characteristics such as the type of facilities and heating systems.
However, the available variables have become less and less relevant to professional users, including
18 years). A relatively simple improvement for the future would thus be to collect age by single years wherever
possible (as part of improved demographic details) and to define some dedicated age categories where necessary.
45
Including the ‘1st
Commission strategy on LGBTIQ equality in the EU’ (press release IP/20/2068), the ‘EU Anti-
racism Action Plan 2020-2025’ (press release IP/20/1654), and the ‘Strategy for the Rights of Persons with Disabilities
2021-2030’ (press release IP/21/810).
46
Responses to the 2nd
Eurostat monitoring survey with NSIs on implementation of the 2021 census round (state
September 2021).
47
Responses to NSI survey (Q4.3.3).
48
DG JUST (2017) Analysis and comparative review of equality data collection practices in the European Union –
Data collection in the field of ethnicity; DG JUST (2017) Analysis and comparative review of equality data collection
practices in the European Union – Data collection in relation to LGBTI people.
49
DG JUST (2016) European handbook on equality data, Section 2.2.4 on categorisation.
50
ESSnet KOMUSO (2018) Quality Guidelines for Frames in Social Statistics.
20
policy users. For instance, housing is the second-lowest-rated data topic (after equality data)
according to professional users among OPC respondents (see Figure 6; only a quarter of the
respondents that expressed an opinion finds the current data sufficient). The consultation of
Commission departments also identified a need for more frequent and timelier housing data,
including appropriate variables to capture energy efficiency of housing and other aspects relevant to
Green Deal monitoring. Relevant European statistics serving these needs will be important to
measure whether the EU remains on the right track towards its ambitious greenhouse gas reduction
goals over the next decades.
During technical consultations with Commission informal expert groups, Eurostat collected several
opinions from NSI experts indicating that housing data are indeed often available from
administrative sources within the countries. In principle, this would lead to cost-efficient production
of more frequent and more relevant housing data, at least in some Member States. The Census
Regulation covers the need for data supporting ‘the protection of the environment and the
promotion of energy efficiency’ in recital 2. However, its legal structure is not flexible enough to
change or add variables in response to such needs (see problem 4). Moreover, the intention to
address this under the present initiative has been subject to more critical opinions at higher levels of
the ESS governance, where the main concern was rather whether these data should remain within
the scope of ESOP, especially when buildings in general (beyond housing) are addressed.
Geographic granularity
Currently, mandatory annual statistics on demography and international migration are mostly only
collected at national level, with some less detailed demographic breakdowns at regional level
(NUTS 2, 3)51
. More detailed regional breakdowns of annual statistics are currently only provided
through voluntary data collections, with the typical consequences outlined in Section 2.2.1. This
annual picture is complemented by mandatory decennial population and housing census outputs
with very detailed breakdowns, including at local administrative unit (LAU) level and – for the
2021 round – some key population indicators georeferenced to a pan-European 1 km square grid.
The evaluation and stakeholder consultation found that most key users of the data believe the
current situation is inadequate, due to a complete lack of subnational data on migration (stocks and
flows). For instance, insufficient geographic detail was the second most frequent critical gap (after
insufficient detail of characteristics) noted for any statistical topic by OPC respondents across all
stakeholder groups52
. Consequently, most OPC respondents across almost all stakeholder groups
(except statistics producers) also put a high priority for a future initiative on adding (mandatory)
regional detail to annual demographic and migration statistics, and on providing functional
typologies (DEGURBA, cities, FUAs) and 1 km square grid data annually53
. This is in line with
findings from the targeted consultation of Commission departments identifying policy needs for
more annual NUTS 3 data, functional typologies, including city and FUA data) and prominently
1 km square grid data54
.
The 1 km square grid data play key roles, both technically and in terms of policy. From a technical
view, the total population at 1 km square grid level is needed to construct the functional typologies
51
Regulation (EC) No 1059/2003 of the European Parliament and of the Council of 26 May 2003 on the establishment
of a common classification of territorial units for statistics (NUTS) (OJ L 154, 21.6.2003, p. 1).
52
Responses to OPC survey (Q2.4).
53
Responses to OPC survey (Q3.3-6).
54
Supporting e.g. monitoring and policies on regional cohesion, urban/rural integration, access to services.
21
(DEGURBA, cities, FUAs), which Regulation (EC) No 1059/2003 defines as clusters of LAUs
according to certain population density and movement patterns based on the grid. This means that
even if Eurostat does not publish European grid data, NSIs need grid information at national level
to determine and maintain the functional typologies. In terms of policy, only 1 km square (or more
detailed) data allow accurate analysis on access to services (e.g. driving time to next hospital), local
crisis management (e.g. locally contained natural disasters like flooding or wildfire) or cross-border
patterns (e.g. functional urban areas clustering in a river basin, where the river happens to mark a
national border). Finally, the baseline includes substantial investment across the ESS to create
infrastructure for the first mandatory grid data collection from the 2021 EU census round55
.
Urban/rural integration
Even though increased availability of regional data at NUTS 3 is widely acknowledged by users,
the angle from an urban/rural dichotomy is often underexposed. This is because NUTS 3 regions
are often still too coarse to capture urban/rural patterns according to functional characteristics
accurately (mostly related to local population density). Therefore, during the targeted consultation
with Commission departments, policy needs were raised for cohesion and the integration of urban
and rural areas, notably for the Commission’s long-term vision for rural areas56
. These policy areas
require any relevant population data cross-tabulated against the degree of urbanisation
(DEGURBA) classification defined in Regulation (EC) No 1059/2003 based on clusters of local
administrative units (LAUs). In addition, population data on cities and their functional urban areas
(FUA) are needed to shed further light on urban/rural patterns.
While European statistics cross-tabulated against DEGURBA have become more and more
prevalent over the past years, no statistics published under the current legal framework have
provided DEGURBA so far. Moreover, annual statistics on the population of cities and FUAs are
currently published regularly on a voluntary basis and outside the ESOP data collections.
Therefore, these data suffer from a lack of harmonisation with ESOP data, in addition to the
identified drawbacks of voluntary collections (Section 2.2.1). Consequently, Figure 6 shows that
urban (47%) and rural (44%) populations also rank among the lowest scoring data topics among
OPC respondents from professional user groups when asked about their agreement on whether the
current data are sufficient.
2.2.4. 2.2.4. Problem 4: The current legal framework is not flexible enough to
adapt to evolving policy needs and to enable using data from administrative
and other new sources in Member States and at EU level
Currently, with regard to the output of the statistical production process, the legal framework sets
out a fixed set of statistical units, variables/breakdowns and cross-tabulations to be produced
regularly, without providing for specific mechanisms to update this statistical content efficiently.
The evaluation has shown that this current static framework has been losing relevance rather
quickly, starting during its implementation period and continuing to the present, due to a framework
that is too rigid and lacking flexibility to adapt data collections to evolving needs. This is confirmed
by the OPC, where a majority of respondents in all stakeholder groups (except statistics producers)
agree only ‘somewhat’ that the legislation is fit for purpose. The Census Regulation represents a
minor exception, as it leaves some room to specify statistical needs for each EU census round
before the census year, thus maintaining a higher relevance of census outputs over time. However,
55
Eurostat (2018) Selected Census 2021 topics on a European 1km2
grid – Cost-effectiveness analysis.
56
COM(2021) 345.
22
the 2021 round has shown that this flexibility was not sufficient in terms of introducing
georeferenced/grid data, which necessitated an ad hoc legal act57
to ensure EU-level completeness,
comparability and coherence.
With regard to input, all three base regulations contain enabling provisions allowing reporting
countries to choose appropriate sources for the statistics, with the Census and Migration
Regulations mentioning administrative sources or registers explicitly. This is in line with a general
principle that has been prevalent in the ESS of ensuring that European statistics are ‘output
oriented’, i.e. minimising legal constraints on inputs and processing of the statistical production
chain58
. Currently, there is a strong ongoing trend towards register-based production systems
drawing from administrative sources. The evaluation has clearly shown that this leads to increased
data availability at significantly reduced production costs. For example, according to the baseline
cost assessment, the median cost of a traditional census across EU-27 countries was roughly 20
times the median cost of a register-based census both in the 2001 and the 2011 census rounds. The
single most significant cost driver in this context for NSIs conducting traditional censuses is the full
field enumeration of the entire population, which entails a complex organisation and coordination at
national level, extensive procedures, and repetitive training and temporary employment of large
workforces of enumerators. However, NSIs that base their production system on administrative
information, for instance from administrative population registers, can usually reduce costs and
avoid such difficulties. Nevertheless, this creates new constraints, as the efficiency of such systems
then depends on the information already available from such sources. If other information is
needed, it must then be modelled (a typical concern being quality) or collected again from field
surveys (which means costs scale up again quickly).
The benefits of moving towards using more administrative and other sources are most pronounced
in combination with a statistical population register as a central processing element of an integrated
multisource statistical production system. According to the targeted NSI survey, 12 Member States
(Austria, Belgium, Bulgaria, Denmark, Estonia, Finland, Italy, Latvia, Netherlands, Slovakia,
Spain, Sweden) already have a statistical population register and 8 more are planning one (Croatia,
Cyprus, France, Greece, Hungary, Lithuania, Malta, Portugal). The main reasons for not having
one, or not having one yet, indicated by the 17 NSIs concerned are the current national legal
framework (mentioned seven times) followed by historical reasons (six)59
. A collaborative network
of ESS experts on population and household frames for social statistics also acknowledged the
significant potential added value of statistical population registers, especially when linked to rich
information from various sources60
.
However, the current statistical legislation on data sources does not allow statistics producers to
easily access appropriate sources held by other owners, including administrative sources. General
access enabling legislation exists in Article 17a of Regulation (EC) No 223/2009 and also very
57
Commission Implementing Regulation (EU) 2018/1799 of 21 November 2018 on the establishment of a temporary
direct statistical action for the dissemination of selected topics of the 2021 population and housing census geocoded to a
1 km2
grid (OJ L 296, 22.11.2018, p. 19); it was adopted as a one-off under Article 14(2) of Regulation (EC)
No 223/2009.
58
e.g. the European Statistics Code of Practice (footnote 3) puts a strong emphasis on outputs.
59
Responses to NSI survey (Q4.1).
60
ESSnet KOMUSO (2019) Quality Guidelines on Frames for Social Statistics (QGFSS) version 1.51 supported by the
DSS (minutes on item 3.4 of the December 2019 meeting); see in particular guideline 3.4 on ‘rich frames’ representing
a type of statistical population register and Annex III on requirements for frame contents (person and household
variables).
23
often in national legislation, as 26 of 29 NSIs confirmed in the survey61
. Such enabling legislation
is generally covered by the GDPR62
and EUDPR63
providing for the processing of personal data for
statistical purposes. However, legal access problems often persist, mainly where specific legislation
on a given source database has more restrictive access limitations. Eurostat is experiencing such
problems with administrative sources held by eu-LISA64
, and 15 NSIs indicate similar issues by
stating that additional legal acts are needed to access some or all specific sources relevant for
population statistics65
. This suggests that the situation could be improved by more explicit sector-
specific rules granting access to statistical offices as needed to all relevant sources for population
statistics. Nevertheless, the evaluation also showed that the effectiveness and efficiency of practical
cooperation agreements between NSIs and administrative source owners are as important as legal
enablers.
Finally, the current legal framework does not encourage the use of new sources, such as privately
held data, including, for instance, geospatial systems or mobile operator data. While the
Demography and Migration Regulations at least allow the use of such sources in principle, the
Census Regulation contains a closed list of eligible source types, excluding any new sources66
. On
top of the issues around access enabling legislation outlined in the previous paragraph, this puts a
strict legal barrier on the use of such new sources for European population statistics. Therefore, the
status quo falls critically short of the state of the art both methodologically67
and legally. Regarding
the legal perspective, the recent Commission proposal on a Data Act (footnote 21) provides
enablers for B2G data sharing, including for official statistics. In this context, the previous
paragraph pointed out that explicit sector-specific enablers (aligned with the Data Act proposal) on
relevant new sources for population statistics may help improve new ESOP legislation in terms of
continued relevance, effectiveness and efficiency.
2.3. 2.3. How likely are the problems to persist?
The evaluation has identified the following four key legislative drivers for the problems described
in Section 2.2.
 Only mandatory data collections with common rules can ensure completeness and
timeliness of statistics at EU level; regulating voluntary data collections that already have
high completeness may lead to significant effectiveness and efficiency gains, as
61
Responses to NSI survey (Q2.8).
62
Regulation (EU) 2016/679 of the European Parliament and of the Council of 27 April 2016 on the protection of
natural persons with regard to the processing of personal data and on the free movement of such data, and repealing
Directive 95/46/EC (General Data Protection Regulation) (OJ L 119, 4.5.2016, p. 1).
63
Regulation (EU) 2018/1725 of the European Parliament and of the Council of 23 October 2018 on the protection of
natural persons with regard to the processing of personal data by the Union institutions, bodies, offices and agencies
and on the free movement of such data, and repealing Regulation (EC) No 45/2001 and Decision No 1247/2002/EC
(OJ L 295, 21.11.2018, p. 39).
64
European Union Agency for the Operational Management of Large-Scale IT Systems in the Area of Freedom,
Security and Justice (eu-LISA).
65
Responses to NSI survey (Q2.8).
66
Article 7 of the Demography Regulation does not list any specific source types and Article 9(1) of the Migration
Regulation allows any ‘other appropriate sources’. On the other hand, Article 4(1) of the Census Regulation lists as
eligible source types ‘conventional censuses’, ‘register-based censuses’ (without a definition), or a combination thereof,
possibly complemented by sample surveys.
67
Eurostat is active in research in this respect; see e.g. Ricciato et al. (2020) Towards a methodological framework for
estimating present population density from mobile network operator data. Pervasive and Mobile Computing 68,
October 20, 101263.
24
considerable EU added value can be generated at limited incremental costs (linked to
problems 1, 2, 3).
 Voluntary data collections are appropriate instruments to pilot the production of new
topics or characteristics, and to foster the incremental capability of national statistical
systems to provide such new data. However, they tend to become inefficient over time, as
recurrent production costs eventually fail to generate substantial EU added value in terms of
completeness and timeliness for all Member States (linked to problems 1, 2).
 Loose legal definitions of statistical concepts lead to a gradual loss of control over
conceptual harmonisation, and thus ultimately over coherence and comparability, over time.
The population base example shows how a default clause originally introduced as an
exception with limited scope turned into a new standard that is detrimental to European
statistics. However, a stronger stance on definitions also requires closer guidance and
monitoring of implementation to ensure real harmonisation (linked to problem 1).
 A legal framework that is too rigid prevents it remaining relevant over time. The current
framework has been losing relevance rather quickly, starting during its implementation and
continuing to the present day. This is due to a lack of flexibility mechanisms for adapting
data collections to evolving needs or for benefiting from opportunities driven by the
availability of new data sources (linked to problems 1-4).
The baseline assumes continuing the status quo regarding the availability of European population
data (see section 5.1). The insight set out above – that the problems emerge directly from legal
constraints enshrined in the current framework – leads to the conservative scenario in which the
current framework will continue losing relevance as observed in the evaluation. This typically leads
to professional users turning to other national or unofficial sources of statistics, both affected by
reduced quality (measured from an EU perspective) and thus leading to poorer policy decisions. A
more optimistic scenario would have to operate on the assumption that evolving policy needs for
data would be sufficiently addressed in the future by voluntary data collections. However, the
experience over the past two decades does not leave much room for such optimism.
3. 3. WHY SHOULD THE EU ACT?
3.1. 3.1. Legal basis
The legal basis for EU intervention in the area of European statistics is Article 338 of the Treaty
on the Functioning of the European Union (TFEU), which confers on the European Parliament
and the Council the power to adopt measures for the production of statistics where necessary for the
performance of the activities of the EU, in line with the ordinary legislative procedure. Article 338
of the TFEU sets out the requirements for producing European statistics, stating that they must
conform to standards of impartiality, reliability, objectivity, scientific independence, cost-
effectiveness and statistical confidentiality. This section assesses whether a revision of the current
legal framework governing European statistics on population is appropriate and justified, in view of
its purpose to ensure high-quality EU population data in line with the statistical principles and
quality criteria applicable to European statistics.
Population data are currently collected under three separate Regulations that neither allow
flexibility nor respond sufficiently to new and emerging statistical needs of users and a recognised
necessity for modernisation (Section 2.3). A partial revision of the legal framework on population
statistics is inevitable for current Regulation (EU) No 1260/2013 on European demographic
statistics, as, under Article 12, it will cease to apply in 2028. A minimal revision is therefore
25
necessary even if the status quo is to be maintained. However, this also generates an opportunity for
a wider review to modernise the broader legal framework for population statistics, which also
includes Regulation (EC) No 763/2008 on population and housing censuses as well as Article 3 of
Regulation (EC) No 862/2007 on statistics on international migration. This would also enable
maximising the added value of capacity already developed for the one-off Commission
Implementing Regulation (EU) 2018/1799 on georeferenced census 2021 outputs, and fully
accounting for the integrated approach to statistical quality set out in Regulation (EC) No 223/2009
on European Statistics.
Article 5 of the Treaty on European Union (TEU) enshrines three principles: conferral,
proportionality and subsidiarity. As an expression of the principle of conferral, Article 338 of the
TFEU empowers the EU to set up European statistics. Furthermore, Article 338(2) of the TFEU
mentions six general principles that EU statistics must follow, namely impartiality, reliability,
objectivity, professional independence, cost-effectiveness, and statistical confidentiality. These
principles are specified in Article 2(a) to (f) of Regulation (EC) No 223/2009 (footnote 2) and
further specified in the European statistics Code of Practice (footnote 3) maintained by the ESS
Committee in line with Article 11 of the same regulation. Article 338(2) of the TFEU also
stipulates that EU statistics must not entail excessive burdens on businesses. This provision,
together with Article 338(1) of the TFEU, according to which statistics must only be produced
where necessary, reflect an expression of the principle of proportionality. Finally, legislative
action on European population statistics falls under supporting competences, where the principle of
subsidiarity authorises an EU intervention only if a specific issue cannot be addressed by the
individual action of Member States (necessity test, Section 3.2) and provided the objective can be
better achieved ‘by reason of the scale and effects of the proposed action’ at EU level (EU added
value, Section 3.3).
3.2. 3.2. Subsidiarity: Necessity of EU action
The ESS provides an infrastructure for statistical information. The system is designed to meet the
needs of multiple users for decision-making in democratic societies. The proposal for this
regulation has been drafted to protect core activities of ESS partners while better ensuring the
quality and comparability of statistics. One of the main criteria that statistical data must meet is to
be consistent and comparable. Member States cannot achieve the necessary consistency and
comparability without a clear European framework, that is to say without EU legislation laying
down the common statistical concepts.
The principle of proportionality aims at identifying the best level of governance to ensure that
decisions meet the public’s needs to the greatest extent possible. Together with the principle of
subsidiarity, the principle of proportionality regulates how the EU exercises the powers conferred
by the Member States on the EU. According to the principle of proportionality, EU action must be
limited to what is necessary to achieve the objectives of the Treaties. It requires the EU to only take
action that is deemed essential to achieve the aim pursued. To ensure compliance with the principle
of proportionality, any EU intervention must ensure a higher level of coherence and comparability
of population statistics across Member States. Further action at EU level is justified in light of the
variety of measures adopted at national level and given that a robust legal framework for the
collection of population statistics is essential for maintaining relevant and comparable statistics at
EU level based on harmonised concepts and approaches to methodology.
Proportionality also requires that the intervention matches the size and nature of the EU-related
problems identified, including the right choice of policy instruments to address the problems. From
this perspective, a legislative solution would be necessary and proportionate, as the problems
26
identified have a genuine EU-wide scope clearly linked to gaps in the current EU legislation
(Section 2). Without further EU legislative action, these problems will persist or worsen. The
current EU legislation will likely continue to become less effective and efficient in achieving its
objectives as many Member States continue to modernise nationally by setting up statistical
population registers and harnessing new data sources. Relevance will also likely decrease further,
as the EU-level statistics are expected to diverge further from users’ needs in terms of content,
desired frequency or timeliness. Without EU legislative action, national approaches will diverge
further, leading to less comparable statistics, which in turn risks compromising policymaking at EU
level.
3.3. 3.3. Subsidiarity: Added value of EU action
The added value of complete and comparable population and demographic statistics at EU level lies
primarily in their significant contribution to various institutional needs and policy areas of the EU
that are highly relevant for many political priorities of the Commission (i.e. An economy that works
for people, Promoting our European way of life, A new push for European democracy). Population
and demographic statistics are also needed to feed into various EU institutional tasks and
procedures laid down by the Treaties, such as national population weights to determine the 65% EU
population quota for qualified majority voting of the Council (Article 16 of the TEU), EU long-
term economic and budgetary projections within the European Semester (Article 121(6) of the
TFEU detailed in Regulation (EU) No 1175/2011), and monitoring of the annual EU demographic
situation (Article 159 of the TFEU).
These data inform EU policies that fall under shared competences (e.g. social policy; economic,
social and territorial cohesion; and the area of freedom, security and justice) and supporting
competences (e.g. health, youth, civil protection and administrative cooperation). Population
statistics are the backbone of other European statistics (sample surveys, national accounts) and used
to calculate per capita indicators. Finally, population and demographic statistics are also designed
to meet the needs of multiple users, for decision-making at all levels in the EU, as well as research
and informing the general public. The EU may therefore adopt measures in this area in line with the
principle of subsidiarity under Article 5 of the TEU.
4. 4. OBJECTIVES: WHAT IS TO BE ACHIEVED?
Figure 8 presents a problem tree showing the logical links between problem drivers and general and
specific problems identified during the evaluation and the consequences of these problems for
policymakers and other data users. As shown in the figure, the regulatory problem drivers reveal
the limitations in the current legal framework, which is not harmonised and lacking in coherence
and flexibility, while at the same time practices in Member States are evolving and societal trends
changing. This leads to specific problems in comparability of population data, data not being as
timely and frequently produced as possible, limitations in the depth and comprehensiveness of data
topics, and quality-related problems. Combined, these problems lead to content and quality gaps
compared to policymakers and users’ evolving needs regarding demographic and societal
challenges.
The general and specific objectives for an EU intervention to modernise European population
statistics have been linked to the problems and their drivers. These are shown in the lower half of
Figure 8 and detailed below.
27
4.1. 4.1. General objective
As shown, the general objective of EU intervention in this case is to better respond to users’ needs
over time and to modernise and enhance the relevance, harmonisation and coherence of European
population statistics.
4.2. 4.2. Specific objectives
Statistics are assessed in the context of the ESS framework for statistical quality68
, and the key
dimensions for this initiative are ‘relevance’, including ‘completeness’, ‘coherence’,
‘comparability’, ‘timeliness’ and ‘frequency’ (or ‘periodicity’). Therefore, the general objective can
be broken down into four specific objectives (SOs) along statistical quality dimensions.
 SO1. Ensure complete, coherent and comparable European population statistics. This means
all European statistics should be available from all Member States and include EU-level
aggregate information. These statistics should follow the same underlying concepts
operationalised in a comparable way, so that the information is comparable across Member
States and coherent across statistical products.
 SO2. Ensure availability of timely and frequent population statistics to meet users’ needs.
This means all European statistics should be provided by all Member States and ready for
publication according to a strict agreed deadline that takes into account user needs for the
time span between reference period and publication of the statistics (timeliness). Statistical
products should also be published with a sufficient granularity of publication dates
considering current user needs (frequency).
 SO3. Provide statistics that are sufficiently comprehensive in terms of relevant topics and
sufficiently detailed in terms of characteristics and breakdowns. This means the European
statistics should address the topical and detail needs of key statistics users and give
information that benefits society at large. Statistics must provide the data evidence needed
for better policymaking in priority areas of the EU as well as support decision-making at
other governance levels or provide the data needed for relevant research or for public
debates.
 SO4. Promote legal and data collection frameworks that are sufficiently flexible to adapt
datasets to evolving policy needs and to opportunities emerging from new sources. This
means the EU statistical legislation should be better able to maintain the relevance of
statistical products over time, by enabling the use of more diverse sources and by offering
mechanisms to update the statistical products. As worked out in Section 2.3, a key issue of
the current legislation is its inability to adapt to evolving input and output contexts. The
evaluation measured this in terms of the gradual loss of relevance of the statistical products.
These specific objectives are SMART69
in that they directly link to the four gaps quantified in
the evaluation (footnote 7). The evaluation provides quantitative indicators on each specific
objective to follow up in the future, for instance over the first one or two production cycles
(including censuses) after adoption of a new legal framework. This completes the pathway from
the problem drivers to the proposed solutions and builds the foundation for developing the
policy measures and options outlined in Section 5.
68
See European statistics Code of Practice (footnote 3) and European statistical system handbook for quality and
metadata reports (2020 edition): https://ec.europa.eu/eurostat/web/products-manuals-and-guidelines/-/ks-gq-19-006.
69
Specific, measurable, achievable, relevant and time-bound in line with the Commission’s Better Regulation Tool #15.
28
Figure 8 – Problem tree (contextual factors)
29
In general, as explained above, the specific objectives are clearly problem-driven as they match the
specific problems worked out in detail in the evaluation (footnote 7) and described in Section 2.
From this perspective, the specific objectives have clear support from essentially all stakeholder
groups because the underlying problems are flagged by those groups. For instance, in the OPC70
,
completeness and comparability (SO1) and insufficient detail (SO3) were mentioned as critical
gaps in the current statistics by significant shares of respondents across all stakeholder groups for
all statistical domains (except marriages and divorces), ranging from 53 respondents on acquisition
or loss of citizenship to 94 on international migration (out of 172 respondents overall). Similarly, a
lack of timeliness and frequency (SO2) were noted as critical gaps by between 19 respondents
(acquisition or loss of citizenship) and 43 respondents (population census outputs). These current
gaps and their relevance to EU policies were expressed even more strongly in the targeted survey of
Commission departments. NSIs generally also acknowledged these gaps across the consultation
activities (OPC, targeted survey and workshops), but typically saw them as less critical than did
stakeholder groups representing statistics users. SO4 emerges as a foresight requirement for a new
legal framework to minimise the risk of accumulating such critical gaps again quickly in the future.
Furthermore, during interviews conducted in the external support study71
, many of the stakeholders
were asked to provide their views on the specific and operational objectives proposed for the ESOP
initiative. Data users and international partners of Eurostat largely supported the policy objectives
shown in Figure 8, though different stakeholders put greater or lesser emphasis or priority on
different aspects. Only a couple of the NSIs commented on the policy objectives and highlighted
some of the practical difficulties of addressing them. Stakeholders interviewed from NSIs and
international organisations also pointed to some of the trade-offs between the policy objectives. For
example, there is a trade-off between ensuring the timeliness and the quality/accuracy of the data.
There may also be a trade-off between adapting the current system and ensuring consistency in the
time series (though some international partners indicated that it can be relatively easy to
retrospectively adjust historic time series). Some NSI stakeholders further mentioned potential
challenges related to personal data protection, particularly if data are to be disaggregated by very
small areas and/or specific population groups, though it is also understood that methodologies are
available or under development that could help to overcome these challenges.
5. 5. WHAT ARE THE AVAILABLE POLICY OPTIONS?
5.1. 5.1. What is the baseline from which options are assessed?
Under the baseline scenario as analysed in depth in the evaluation, each statistical domain will
continue to be implemented based on the three separate legal acts currently in force (Census,
Demography and Migration Regulations). One issue is that the Demography Regulation contains a
‘sunset clause’ stating that it will cease to apply on 31 August 2028. However, mandatory data
collections under that Regulation are currently generally running well with no reason to assume that
this status quo would be seriously questioned by relevant stakeholder groups (including statistics
producers) in the baseline scenario. It is therefore assumed at the baseline that the Demography
Regulation would simply be amended to extend its application (see Section 5.3). In summary, the
baseline has the characteristics described below.
70
Responses to OPC survey (Q2.4).
71
ICF (2022) Study supporting the evaluation and impact assessment of European statistics on population – final report
on impact assessment support.
30
There are no changes regarding harmonisation of statistics across Member States. The limited
harmonisation of the population base definitions remains, which would likely mean that the EU-
level comparability of national definitions will continue to diverge further, as observed in the
evaluation, in the wake of an ongoing and still increasing transition to national register-based
systems in most Member States.
Demography, international migration, and population and housing census statistics are implemented
as separate statistical processes. Member States may or may not update, streamline or integrate
their national processes in future, according to their own considerations. Eurostat keeps the existing
data flows from Member States unchanged. This means possible efficiency gains by embracing
administrative and other new sources remain uncoordinated at EU level and up to Member States.
There are no new or more detailed statistical outputs adding EU value, as the mandatory data
requirements are not updated. Voluntary data collections when agreed with Member States lead to
gaps in terms of data completeness and timeliness and cannot sufficiently deliver EU value, as
shown in the evaluation.
The legal framework is not flexible enough to adapt the mandatory data collections to evolving
policy or data environments. This means that, as needs evolve, even if new voluntary collections
should perform better than the baseline assumes, there are no mechanisms to reap the full benefits
at EU level and in terms of the efficiency of such achievements. With regard to evolving data
environments, the current legislation generally does not sufficiently enable the use of new sources
(e.g. B2G), and the Census Regulation currently even excludes such uses.
5.2. 5.2. Description of the policy options
The aim of any policy option beyond the baseline is to deliver on each of the specific objectives set
out in Section 4.2. To achieve this, more detailed policy measures were identified under 17 topics,
where each topic acts on exactly one specific objective. More precisely, each topic provides a set of
different policy measures typically relating to varying levels of ambition for that topic (see
Table 2). Each overarching policy option is then constructed bottom-up as a combination of policy
measures from all 17 topics, thus acting with varying ambition on all specific objectives.
Before grouping the various measures into policy options, however, a first feasibility assessment of
each individual topic and measure was carried out. This process involved assessing how the topics
and measures relate to one another (mutually exclusive or complementary), and carrying out an
initial screening of their likely effectiveness, as well as their technical, operational, economic and
legal feasibility. This initial screening had the results described below.
 A topic on the census frequency (original measures to maintain status quo, split the census
into decennial and annual outputs depending on the relevance of the data, and annual
census) has been discarded because combinations of other measures factually imply that the
frequency of (parts of) the census changes by default.
 Two of the measures relating to the topic on output flexibility of the legal framework have
also been discarded. One measure ‘Include limited flexibility to adapt statistics; use
voluntary data collections to serve emerging needs’ was discarded because it was assessed
as being very similar to the status quo. The other measure ‘Include effective mechanisms to
adapt statistics more quickly and efficiently when there is an added value for the EU, even if
the administrative cost and burden are high’ was discarded as unrealistic.
The remaining list of the most feasible individual measures, which have been retained for further
analysis, is presented in Table 2.
31
Policy options were then constructed by grouping measures based on four main characteristics:
 harmonisation of statistics where the main focus is on the population base definition
(status quo, harmonisation with justified exceptions, strict harmonisation);
 integration of statistical processes (status quo, improved statistical processes, integrated
statistical processes);
 statistical outputs (status quo, limited upgrade, more expansive upgrade, major upgrade);
 framework flexibility (status quo, limited flexibility, effective mechanisms).
Table 2 lists all policy topics and potential measures with a short description of each. Figure 9
shows the resulting grouping of the measures into seven policy options:
A – Baseline as described in Section 5.1.
Policy options B.1 and B.2 – For these options, there is some simplification of EU-level data
collections, but the statistical production processes remain separate and somewhat improved
depending on plans of Member States, and the harmonisation of the definition of population
remains limited. However, these options introduce two incremental levels of ambition in terms of
statistical outputs (limited upgrade for B.1, more expansive upgrade as well as improved flexibility
for B.2).
Policy options C.1 and C.2 – In these options, EU data collections are simplified and the statistical
production processes remain separate and somewhat improved depending on plans of Member
States, but the harmonisation of the population base definition is improved relative to policy
options A and B. Policy options C.1 and C.2 introduce the same two incremental levels of ambition
in terms of statistical outputs as policy options B.1 and B.2 described above (limited upgrade for
C.1, more expansive upgrade as well as improved flexibility for C.2).
Policy options D.1 and D.2 – These are the most ambitious options, introducing strict
harmonisation of the population base definition as well as a major upgrade of statistical outputs.
These options also include effective flexibility mechanisms to adapt statistics more quickly and
efficiently to meet emerging user needs and exploit new data sources or methods. While option D.1
leaves statistical production processes within the competence of Member States, option D.2
requires all Member States to set up statistical population registers, thus initiating the
redevelopment and integration of statistical processes at national and EU levels.
From the perspective of the four main characteristics introduced above, these policy options thus
implement incremental ambitions to improve as follows (see also Table 1).
With regard to the harmonisation of population base definitions, the status quo is maintained in
policy options B (both B.1 and B.2). Incremental levels of harmonisation would be introduced for
policy options C and D as follows. In policy option C (both C.1 and C.2), a harmonised population
definition would be introduced for all datasets, but Member States would still be able to use
national population definitions in justified exceptions with limited impact on comparability of
statistics across Member States. Policy option D (both D.1 and D.2) introduces a strictly
harmonised population definition whereby Member States would be required to use the strict usual
residence concept including the 12-months rule to define their populations for all datasets for
European statistics purposes.
On statistical processes, all options will revise the legal framework and thus offer an opportunity
to streamline it, which would lead to simplified data collection procedures between NSIs and
Eurostat. In the most ambitious option (D.2), all Member States would additionally be required to
32
set up a statistical population register. This would align with earlier ESS expert recommendations
in the context of population and household frames for social statistics (footnote 60).
In terms of statistical outputs, three incremental levels of ambition are introduced across the
options as follows. In policy options B.1 and C.1, the annual statistical outputs for demography and
migration would become more detailed to a limited extent, either in terms of characteristics or
geographical disaggregation. In policy options B.2 and C.2, statistical outputs on demography and
migration would become even more detailed compared to policy options B.1 and C.1. Some
currently voluntary datasets and breakdowns, as well as proportionate time series updates, would
become mandatory. In all policy options B and C, the timeliness would be improved. In policy
option D (both D.1 and D.2), the required statistical outputs on migration and demography as well
as regulation of currently voluntary datasets and breakdowns go further than in policy options B.2
and C.2. Moreover, timeliness would be improved and the NSIs would be required to coordinate
their census data releases. LAU-level data would become annual, and time series would be updated
entirely.
On the flexibility of the legal framework governing European statistics on population, the
policy options B-C introduce limited flexibility to adapt statistics to emerging needs by piloting
voluntary collections, combined with a legal mechanism to regulate successful pilots and strengthen
the legal base to benefit from administrative and other new data sources. Moreover, policy options
B.2 and C.2 introduce basic interoperability of statistical population registers. In policy options D.1
and D.2, the flexibility of the legal framework would be substantially improved by setting up full
EU-wide interoperability of statistical population registers as well as an effective mechanism to
adapt statistics more quickly and efficiently, as proportionate, to cover more general emerging
needs.
Table 1 – Comparative ambition of the policy options going beyond the baseline (A) regarding the four main characteristics
discussed in the text (scale: no ambition ‘0’, otherwise ‘+’, ‘++’ or ‘+++’).
Option Harmonisation Process integration Statistical outputs Flexibility
A (baseline) 0 0 0 0
B.1 0 + + +
B.2 0 + ++ ++
C.1 ++ + + +
C.2 ++ + ++ ++
D.1 +++ + +++ +++
D.2 +++ +++ +++ +++
33
Table 2 – Detailed overview of policy topics and their policy measures and sub-measures
Policy topic Policy measure Acting on
SO
1 Statistical population
frames or registers
1.1 Maintain status quo -
1.2 NSIs set up statistical population registers; the
registers would cover total population exhaustively
and accurately and enable direct extraction of up-to-
date statistics
1
2 Harmonisation of the
population definition
2.1 Maintain status quo -
2.2 Publish European population statistics based on a
strictly harmonised population definition (based on
usual residence concept) for all datasets
1
2.3 Publish European population statistics based on a
harmonised population definition (based on a usual
residence concept) for all datasets, with justified
exceptions and limited impact on comparability of
statistics across Member States
1
3 Regional detail of
annual population
statistics
3.1 Maintain status quo -
3.2 NSIs provide all annual population data at NUTS 3
level
3
3.3 NSIs provide annual population data by functional
geographic units (cities, functional urban areas) and
typologies (DEGURBA classification)
3
3.4 NSIs provide essential annual population data by
functional geographic units (cities, functional urban
areas) and typologies (DEGURBA classification)
3
4 Regional detail of
annual migration
statistics
4.1 Maintain status quo -
4.2 NSIs provide migrant stocks and migration flows at
NUTS 2 level
3
4.3 NSIs provide migrant stocks and migration flows at
NUTS 3 level
3
5 Statistics at
local/municipality
level
5.1 Maintain status quo -
5.2 NSIs provide LAU data annually 3
6 Georeferenced
statistics
6.1 Maintain status quo -
6.2 NSIs provide georeferenced data every 10 years 3
6.3 NSIs provide georeferenced data annually 3
7 Infra-annual statistics 7.1 Maintain status quo -
7.2 NSIs provide detailed monthly data on population,
births, deaths and migration regularly
2
7.3 NSIs provide less detailed quarterly data on
population, births, deaths and migration regularly
2
8 Timeliness of annual 8.1 Maintain status quo -
34
Policy topic Policy measure Acting on
SO
population statistics
8.2 Quality first: NSIs provide annual data more quickly
as far as possible, while ensuring high statistical
quality before release
2
8.3 Timeliness first: NSIs provide annual data
substantially more quickly, at the expense of revising
the data more often to improve quality gradually
2
9 Timeliness of EU
census outputs
9.1 Maintain status quo -
9.2 NSIs publish EU census data within a shorter legal
deadline but without the obligation to coordinate the
exact release date at EU level
2
10.3 NSIs publish EU census data within a shorter legal
deadline and with the obligation to coordinate the
release date
2
10 Details of migrant
population
characteristics
10.1 Maintain status quo -
10.2 Define migrant groups in more detail (e.g. by single
years of age where possible)
3
10.3 Define migrant groups in more detail (e.g. by single
years of age where possible), including socio-
economic details
3
11 Details of migration
flow characteristics
11.1 Maintain status quo -
11.2 Ensure that statistics on migration flows within the EU
become as detailed on demographic aspects as
statistics on migration flows from/to non-EU countries
3
11.3 Ensure that statistics on migration flows within the EU
and from/to non-EU countries become more detailed
and comprehensively cover demographic (e.g. single
years of age where possible) and socio-economic
aspects
3
12 Voluntary statistics on
marriages, divorces,
legally induced
abortions, infant
mortality and loss of
citizenship
12.1 Maintain status quo -
12.2 Set requirements for all NSIs to provide mandatory
statistics on marriages, divorces, legally induced
abortions, infant mortality and loss of citizenship
3
13 Voluntary breakdowns
of live births, deaths
and international
migration
13.1 Maintain status quo -
13.2 Set requirements for NSIs to provide existing
voluntary breakdowns on a mandatory basis
3
14 Statistics on equality
and non-discrimination
characteristics
14.1 Maintain status quo -
14.2 Provide size and demographic characteristics of the
groups referred to
3
14.3 Provide size and demographic characteristics of the
groups referred to, and disaggregate other statistics by
such group characteristics where relevant
3
35
Policy topic Policy measure Acting on
SO
15 Flexibility of the legal
framework for
statistical outputs
15.1 Maintain status quo -
15.2 Include limited flexibility to adapt statistics; introduce
limited executive power on the Commission to enable
the use of initial voluntary data collections to serve
emerging needs with a mechanism to make them
mandatory later depending on piloting results
4
15.3 Include more effective executive power on the
Commission to adapt statistics more quickly and
efficiently (e.g. to meet emerging user needs, exploit
new data sources or methods), where the
administrative cost and burden are proportionate to the
added value for the EU
4
16 Time series of annual
statistics
16.1 Maintain status quo (update historic time series where
feasible; NSIs choose which time series to update)
-
16.2 Update historic time series proportionately (from 1990
for demography and from 2007 for international
migration) to ensure a certain degree of comparability
over time
1
16.3 Update historic time series entirely (from 1960 for
demography and from 1990 for international
migration) to ensure comparability over time
1
17 Flexibility of the legal
framework for inputs
and new sources
17.1 Maintain status quo -
17.2 Better use of administrative data sources at national
and EU level, including reuse of interoperability
systems
4
17.3 Better use of privately held data (B2G) by effectively
enabling the mechanisms provided for in the Data Act
proposal (footnote 21) for European population
statistics
4
17.4 Basic interoperability of statistical population registers 4
17.5 Full EU-level interoperability of statistical population
registers
4
36
Figure 9 – Mapping of policy options to topic measures (the cell shading represents the different levels of ambition of each sub-measure: the darker, the more ambitious)
37
5.3. 5.3. Options discarded at an early stage
Another option has been taken into consideration, in which demography statistics would become
deregulated because a sunset clause is included in the Demography Regulation. However, this
option has not been included in the shortlist of options for further analysis as it was too similar to
the baseline scenario. The rationale is that this sunset clause states that the Demography Regulation
will cease to apply on 31 August 2028. This impact assessment looks at the impacts over a 10-year
period, until 2031. Therefore, the differences between the baseline scenario and this option would
only apply to the 3 last years of the time period assessed. Moreover, the evaluation has shown that
the currently regulated data collections are mostly running well, so it can be assumed that an
amendment of the Demography Regulation could be put in place simply by extending its
application.
6. 6. WHAT ARE THE IMPACTS OF THE POLICY OPTIONS?
This section provides a partial quantitative assessment of the efficiency of the policy options, with
impacts monetised to the extent possible based on the available evidence. It also contains a
qualitative assessment for each of the policy options covering the following criteria:
 effectiveness, including impacts on fundamental rights and wider environmental and social
impacts;
 coherence with the EU policy objectives, including consistency with the proportionality
principle;
 efficiency.
A detailed analysis of impacts is provided in Annex 2 based on the external support study
(footnote 71).
Scoring approaches used in the report
The consistency of each policy option with the principle of proportionality is rated as ‘Pass’,
‘Uncertain’ or ‘Fail’.
The options have also been scored qualitatively against the criteria of effectiveness and coherence.
Each option was scored against these three criteria using the following scoring system:
Level of impact Score
Very strong negative impact - 4
Strong negative impact - 3
Moderate negative impact - 2
Slight negative impact - 1
No impact compared to baseline 0
Slight positive impact + 1
Moderate positive impact + 2
Strong positive impact + 3
Very strong positive impact + 4
38
However, efficiency has been assessed qualitatively by confronting the incremental costs and
benefits for each policy option.
The baseline sets a ‘0’ against all three criteria (on efficiency, there are no incremental costs or
benefits). All policy options B-D have been scored relative to this baseline using the scoring system
above. The options have also been scored relative to one another. Thus, two or more options with a
similar impact against the same criterion have been given a tied score. For most criteria, it is clear
that one or more options will have a very strong positive (+4), or very strong negative (-4) impact.
For these criteria, the option(s) with the strongest impact received the highest/lowest score (+/-4),
and all other options are scored relative to this maximum and the ‘0’.
For a few criteria (e.g. social impact and impact on fundamental rights), the maximum score is 3, a
strong positive impact. For these criteria, the score of 4 is not reached due to risk discounts related
to concerns expressed by some stakeholders. Thus, a lower score was used as the comparator.
Scores are given based on the evidence gathered from the consultation and desk research.
6.1. 6.1. Quantitative analysis of the options
Quantitative analyses of costs have been carried out for the baseline and each of the six options,
over a 10-year period (2022-2031)72
. The costs associated with the baseline represent an estimation
of the current costs, assuming these costs would remain constant over 10 years. The costs assessed
for policy options B.1, B.2, C.1, C.2, D.1 and D.2 are incremental relative to the baseline (denoted
as A). These only estimate the added costs associated with the implementation of each of the
measures/action required by each option, over and above the baseline situation, i.e. costs that would
be incurred in any scenario. As such, total costs associated with each option should be interpreted
as being equal to the costs reported for the baseline in addition to those reported for every other
relevant option. A specific cost itemisation was made and the different ‘cost items’ were then
matched to the generic cost itemisation in the Better Regulation guidelines (BRG). Table 13 in
Annex 2 (part 1.1) presents the cost types considered in the quantitative assessment of all options
and how these relate to the specific cost items.
Most benefits associated with each option, except for some potential cost savings related to reduced
administrative burden, have not been quantified and are therefore assessed qualitatively in
Section 6.2. A more detailed costs-and-benefits assessment of each option is provided in Annex 2.
6.1.1. 6.1.1. A – Baseline
In the baseline scenario A, the harmonisation of population base definitions will remain limited by
default provisions under the current legislation, statistical production processes will also remain up
to Member States and the current requirements in terms of statistical outputs will continue to be the
same. Finally, the current lack of flexibility to address evolving statistical needs as well as to
benefit from opportunities from new data sources will persist.
The cost assessment of the baseline scenario was carried out based on the evaluation results. For all
policy measures included in the baseline, and as always in a baseline scenario, the status quo will
72
The detailed cost assessment is part of the external support study (footnote 71).
39
remain. It should be noted that even though there is a sunset clause in the Demography Regulation,
it was assumed in the baseline that a minimal amendment would be adopted to extend its
application (see Section 5.3).
All recurrent costs incurred by the Census, Demography, and Migration Regulations were estimated
to remain constant and were calculated over a 10-year period, for both the Commission (Eurostat)
and the 27 Member States and their respective NSIs. However, the baseline scenario is not one of
‘no change’. The baseline is dynamic in the sense that it takes into account ongoing and/or planned
activities and takes into account the effects of national action (both legislative and voluntary) that
can be expected to be taken regardless of any future EU initiative. Therefore, costs and cost savings
associated with maintaining statistical population registers (for those Member States that have
implemented or are planning to implement one regardless of this initiative), implementing a strict
usual residence definition of the population (for those Member States that currently adopt this
definition) or providing voluntary statistics are also included in the dynamic baseline. The rationale
for the including these elements in the baseline is as follows.
 In terms of statistical processes, Member States that already have a statistical population
register73
are assumed to continue to incur costs to keep these registers up to date. These
costs are therefore included in the baseline. Similarly, some NSIs plan to set up a statistical
population register in the near future74
and will therefore incur implementation costs as well
as cost savings related to subsequent efficiency gains, regardless of a potential new EU legal
base. These upcoming effects cannot be attributed to a potential new EU requirement on
NSIs to set up and maintain national statistical registers (see measure 1.2), and are therefore
also included in the baseline75
.
 On the harmonisation of population definitions, some Member States already use a
population base definition based on the strict usual residence concept76
. Following the same
reasoning as above, those Member States already incur related costs and will continue to
incur them. These costs cannot be attributed to a potential new EU requirement to use that
same definition (measures 2.2 or 2.3) and are thus included in the baseline.
 Similarly, some Member States already provide datasets voluntarily. Data currently
collected on a voluntary basis incur costs for both the Commission (Eurostat) and NSIs.
These costs are likely to remain whether a new regulation makes this voluntary data
collection mandatory or not. These costs associated with voluntary data are therefore also
included in the baseline.
73
NSIs survey responses (Q. ‘Does your NSI maintain a national population register for statistical purposes?’ The
following NSIs responded ‘Yes’: Austria, Belgium, Bulgaria, Denmark, Estonia, Finland, Italy, Latvia, Netherlands,
Slovakia, Spain, Sweden).
74
NSIs survey responses (Q. ‘Does your NSI maintain a national population register for statistical purposes?’ The
following NSIs responded ‘No, but planned’: Croatia, Cyprus, France, Hungary, Lithuania, Malta, Portugal; Greece
imputed due to auxiliary information).
75
The contract support study (footnote 71) estimated the cost savings for Member States planning a statistical
population register using the same model as in Section 6.1.3. However, given the non-reciprocity between having a
register-based census and having a statistical population register, it was assumed conservatively that these costs savings
would be equal to one tenth of the annual costs savings indicated by that model.
76
NSI survey responses (Q. ‘Which definition(s) of the population is used by your NSI?’. The following NSIs
responded ‘strict usual residence’: Bulgaria, Croatia, France, Hungary, Ireland, Lithuania, Malta, Portugal, Romania).
40
Table 3 – Overview of baseline total and incremental costs in million 2021 EUR by policy option, and split by one-off and
recurrent per 10 years on Eurostat and on all aggregated 27 Member States.
Option
Eurostat Member States/NSIs (all 27)
One-off
Recurrent (over
10 years)
One-off
Recurrent (over
10 years)
A – baseline total 13.56 1 677.41
B.1 – incremental 0.19 0.10 5.82 6.97
B.2 – incremental 0.39 0.37 17.77 33.98
C.1 – incremental 0.29 0.61 12.46 25.96
C.2 – incremental 0.48 0.89 24.41 52.97
D.1 – incremental 0.58 1.18 41.41 75.60
D.2 – incremental 0.83 1.83 50.42 128.92
Table 4 – Overview of median costs per capita for the three Member State groups (see text) in 2021 EUR per capita.
Option
Member States/NSIs by group (median)
Group 1 Group 2 Group 3
A – baseline total 2.37 4.91 4.07
B.1 – incremental 0.07 0.04 0.02
B.2 – incremental 0.24 0.15 0.06
C.1 – incremental 0.23 0.13 0.05
C.2 – incremental 0.44 0.24 0.07
D.1 – incremental 0.68 0.37 0.09
D.2 – incremental 0.98 0.50 0.19
Table 5 – Overview of average costs per capita in the three Member State groups (see text) in 2021 EUR per capita.
Option
Member States/NSIs by group (average)
Group 1 Group 2 Group 3
A – baseline total 3.00 7.33 8.18
B.1 – incremental 0.10 0.18 0.23
B.2 – incremental 0.35 0.66 0.91
C.1 – incremental 0.37 0.34 0.54
C.2 – incremental 0.63 0.81 1.22
D.1 – incremental 0.90 1.12 2.48
D.2 – incremental 1.22 1.68 3.19
D.2 – incremental 1.22 1.68 3.19
41
In summary, it is thus important to note that the baseline costs estimated as part of this impact
assessment go beyond the costs associated with the current legal framework, in that they also
include current costs incurred on aspects affected by some policy measures and options. As
described in Section 2.2.4 and substantiated by the evaluation findings, there is a clear pattern
expected in the overall baseline costs depending on whether NSIs can use administrative and other
appropriate sources efficiently or not. While it is not straightforward to categorise Member States
under this criterion, the existence (or not) of a statistical population register is a good proxy.
Therefore, Member States are split broadly into three groups as follows:
 group 1 – already have a statistical population register77
;
 group 2 – currently do not have a statistical population register but are planning to set up
one in the near future regardless of any EU intervention78
;
 group 3 – do not have and are not planning a statistical population register79
.
Indeed, the contract support study (footnote 71) has confirmed the expected pattern of higher per
capita costs in groups 2 and 3 than in group 1, both in medians (Table 4) and in averages (Table 5)
across members of each group. Note that the pattern holds despite cost savings included in the
baseline for group 2, likely due to the very conservative approach taken in the estimation model
(footnote 75).
6.1.2. 6.1.2. Incremental costs incurred by policy options B-D
A comprehensive and detailed quantitative assessment of incremental costs incurred by policy
options B-D is provided in Annex 2 (part 1.1). Table 3 shows the aggregated baseline and
incremental costs in 2021 EUR by policy option. The cost model is aligned to the basic
characteristics of the policy options in terms of harmonisation – process integration – upgrade of
statistical outputs (see Section 4.2 and Figure 9 for context). Incremental costs related to flexibility
characteristics are considered indirect and largely depend on unknown context factors (e.g. specific
new data needs or data source environments in the longer-term) so they could not be quantified.
Table 3 also shows that the largest part of both baseline and incremental costs (97.7% to 98.5%
depending on the policy option) are, or would be, incurred by Member States, more precisely by
their national statistical production systems coordinated by the respective NSIs. This reflects the
structure of the ESS governed under Regulation (EC) No 223/2009 (footnote 2), where Member
States produce statistics under the coordination of Eurostat and governance of the ESS Committee
(ESSC). Key features of each policy option regarding incremental costs are summarised below.
B.1 causes incremental costs compared to the baseline solely through the limited upgrade of
statistical outputs, including for instance improved detail and timeliness of statistics and decennial
population grids.
77
Austria, Belgium, Bulgaria, Denmark, Estonia, Finland, Italy, Latvia, Netherlands, Slovakia, Slovenia, Spain and
Sweden following replies to the targeted NSI survey. Note that Slovenia was imputed to group 1 following findings
form the case study in conjunction with a broad definition of ‘statistical population register’.
78
Croatia, Cyprus, France, Greece, Hungary, Lithuania, Malta and Portugal following replies to the targeted NSI
survey (Greece communicated information later).
79
Czechia, Germany, Ireland, Luxembourg, Poland and Romania following replies to the targeted NSI survey (Q.
‘Does your NSI maintain a national population register for statistical purposes?’ These NSIs responded ‘No, and not
planned’).
42
B.2 causes higher incremental costs than B.1 solely through a more expansive upgrade of statistical
outputs, including improved frequency and timeliness, regulating some of the currently voluntary
datasets and introducing annual population and housing grids as well as basic equality data.
C.1 and C.2 cause higher incremental costs than B.1 and B.2 respectively solely through the
improved harmonisation of the population base definition based on the strict usual residence
concept, subject only to justified exceptions (measure 2.3).
D.1 causes higher incremental costs than C.2 by fully harmonising the population base definition
without exceptions and through a major upgrade of statistical outputs, including monthly
demographic data, annual data at LAU level and more detailed equality data.
D.2 causes higher incremental costs than D.1 solely through the obligation on Member States in
group 3 to set up a statistical population register. However, this option also generates long-term
recurrent cost savings for these Member States as described in Section 6.1.3.
Regarding quantitative incremental cost estimates by policy option and Member State group, Table
4 and Table 5 show that average incremental costs per capita are generally increasing from group 1
to group 3, while median incremental costs per capita follow the opposite pattern (decreasing from
group 1 to group 3). This is in contrast to the baseline assessment in Section 6.1.1 (where per capita
cost patterns are consistent for averages and medians over group members), and it suggests that the
cost model provided in the contract support study (footnote 71) was not detailed enough to allow
reliable quantitative estimates broken down by specific Member States or Member State groups.
Nonetheless, the stakeholder consultation and regular expert consultations with NSIs provide three
broad qualitative patterns described below of how adaptation or modernisation challenges vary
across Member States.
1. While upgrading statistical outputs does depend a lot on the specific details of the upgrade
(e.g. whether only timeliness or frequency is improved or whether new variables or
breakdowns are to be added that may or may not be available from sources already in use),
it is generally acknowledged that an integrated statistical production system in place, with a
statistical population register as a backbone, does reduce the costs of such upgrades. This
means that Member States in group 3 (and in group 2 while national efforts to set up a
statistical population register are ongoing) will likely face more severe challenges than
group 1 in adapting to any of the policy options. The same pattern is also relevant when it
comes to preparing for additional flexibility (either to address changing policy needs or to
set up data-sharing systems with new sources).
2. Most of the modernisation activities covered to a varying degree in the policy options would
entail upfront investment (adaptation costs) that are not necessarily proportional to country
size in terms of population or gross domestic product. For instance, upgrading IT
infrastructure or introducing new data products (e.g. infra-annual publications) typically
entails incremental financial and human resources that are at least partly comparable in
volume across countries, whereas smaller countries typically have fewer staff and smaller
budgets for official population statistics. This means that these countries are generally
expected to face relatively more severe challenges than larger countries in adapting to any
of the policy options.
3. The correct implementation of a harmonised population definition based on a strict usual
residence concept as a basis for annual and more frequent statistics is challenging for all
Member States, in particular in combination with the expected upgrading of statistical
outputs. It is generally acknowledged that an integrated statistical production system based
43
on administrative and other readily available sources is needed to produce detailed statistical
outputs frequently and with shorter deadlines. However, multisource production systems
ultimately based on administrative population registers have well-known coverage issues
with respect to the usually resident population. While efficient methods like signs of life are
available in principle to control over-coverage (i.e. to identify persons not factually present
at a given address), ESS80
and international81
guidance generally acknowledges that the only
effective method currently available to check for systematic under-coverage of the registers
against the usually resident population is to carry out dedicated field surveys. The costs of
such field surveys scale with their frequency and geographic area covered and would be
incurred by all Member States using (or transitioning towards) integrated register-based
production systems with a proper implementation of the usual residence concept.
From these points, in particular 1 and 2, it may be roughly summarised that the smallest Member
States in groups 2 and 3 are likely to face relatively the most severe challenges in adapting to any of
the policy options.
6.1.3. 6.1.3. Incremental cost savings (benefits) incurred by policy option D.2
While incremental quantitative costs could not be broken down by specific Member States or
Member State groups due to a lack of granular information, the quantitative evidence for baseline
costs by Member State is more robust and makes it possible in principle to model some specific
cost savings. This is particularly true for census costs, for which there is comparably rich granular
information from the 2001 and 2011 rounds that may be used to model efficiency differences
between production systems, either based on statistical population registers or having to cope
without them. This is used here as a rough proxy to estimate potential cost savings expected
specifically from policy option D.2 that would introduce such statistical registers for all Member
States.
The detailed baseline cost assessment by Member State has shown that there is a significant gap
between Member States producing their annual and census statistics from an up-to-date statistical
population register (group 1 in Section 6.1.1) and Member States currently coping without such a
register (groups 2 and 3). According to the baseline cost assessment, the median cost of a traditional
census across EU-27 countries was roughly 20 times the median cost of a register-based census
both in the 2001 and the 2011 census round. A rough estimation of respective potential cost savings
in the future, accounting for scale effects from changing population sizes, amounts conservatively
to EUR 2 (EUR 1-4 with sensitivity margin of 50-200% around the best estimate) per capita less in
a census round82
, incurred only by six Member States not currently planning to set up a statistical
population register83
(Annex 3). The total potential cost savings at EU level per future census round
could thus amount to EUR 281 million (between EUR 141 million and EUR 563 million
according to sensitivity margin). This is roughly between two and eight times the total combined
one-off and recurrent (over 10 years) incremental cost estimate for all affected Member States to set
up and maintain such registers (see Annex 2 part 1.1.7).
80
ESS Quality Guidelines on Frames for Social Statistics (footnote 60).
81
UN Handbook on Registers-Based Population and Housing Censuses (second draft version presented to the UN
Statistical Commission in March 2022, see footnote 17).
82
External support study (footnote 71).
83
NSIs survey responses (Q. ‘Does your NSI maintain a national population register for statistical purposes?’ The
following NSIs responded ‘No, and not planned’: CZ, DE, IE, LU, PL, RO).
44
Similarly, recurrent cost savings can also be expected for the production of annual and infra-annual
data, even at the highest demographic and geographic details and improved timeliness, due to the
continuous availability of a single statistical source at microdata (individual person record) level
from which the tabulations can be extracted. This would incur a substantial discount on the
incremental costs related to the major upgrade of statistical outputs in policy option D.2, but such a
discount was impossible to quantify reliably on top of the estimate of these costs.
6.2. 6.2. Qualitative analysis of the options
Most benefits associated with each policy option, except for some potential cost savings related to
reduced administrative burden (see Sections 6.1.3 and 8.2), have not been quantified. This is in part
due to a lack of available data on the benefits associated with population statistics, and in part due
to the nature of the benefits that are likely to be generated by this initiative. For many of the
benefits considered, their effects are more indirect and variable across Member States and
stakeholder groups, which make them challenging to quantify or monetise. For example, the
benefits to data users from increased access to high-quality European statistics on population would
depend on several additional factors, such as how these data would be used or the cost of accessing
data through alternative sources. In addition, benefits to policymaking at EU or other governance
levels, from improved access to reliable, comparable population data across the EU, would be
difficult to monetise since these would depend on several contextual factors, such as the types of
policies at each governance level that rely on population data or what the impact of evidence
quality would be.
For these reasons, respective benefits have been considered qualitatively; and Table 6 shows to
what extent each of the policy options B-D achieve incremental benefits compared to the baseline,
by beneficiary groups and specific benefit items. The tables in Annex 2 (part 1.2) provide a detailed
and systematic qualitative assessment of the proportionality and incremental impacts by option,
based on the policy measures included in each option (see Figure 10) and how these affect the
effectiveness, coherence with wider EU objectives and efficiency; and applying the qualitative
scoring method introduced at the beginning of Section 6.
Key features of each policy option regarding incremental policy measures and their benefits on
stakeholder groups are summarised below. Stakeholder groups’ views on each policy option were
collected at three dedicated option validation workshops: one with participants from NSIs, one with
representatives of Commission departments and one with other professional users (institutional
users, researchers, NGOs). Earlier consultation activities could not refer directly to the policy
options as these were not yet finalised at the time, but the OPC did ask for respondents’ priorities
among the various potential measures considered under each of the policy topics (see Table 2).
Mapping their answers to policy options (Figure 9) makes it possible to calculate an OPC-based
relative preference for the policy options by stakeholder group. The result, shown in Figure 10,
corroborates the following summaries by option.
B.1 causes modest incremental benefits from better data evidence for various professional users,
including policymakers (measures 3.2, 4.2, 6.2, 8.2, 9.2, 10.2, 11.2) but fails to deliver on key
aspects like geographic granularity or specific topics of policy interest. Statistics producers,
including Eurostat, benefit moderately from simplified procedures due to a streamlined legal
framework for all datasets and lighter adaptation processes to evolving data needs (measure 15.2).
Source data providers benefit from simplified procedures due to a clearer legal base for sharing
45
relevant data (measures 17.2 and 17.3). During the option validation workshops, participants from
NSIs and professional users (Commission departments, other institutional users and researchers)
identified policy option B.1 as realistic, i.e. feasible to implement84
.
B.2 causes larger incremental benefits for professional users than B.1 in terms of further improved
data evidence for various policy and institutional purposes (additional measures 3.4, 4.3, 6.3, 7.3,
8.3, 10.3, 11.3, 12.2, 13.2, 14.2, 16.2). Statistics producers additionally benefit from the
interoperability of statistical population registers for simplified sharing of relevant data
(measure 17.4). The feedback from the option validation workshops was mixed: a majority of NSI
participants stated that the measures under policy option B.1 would be unrealistic due to the
increased burden involved, while the opposite view was broadly expressed among professional
users (see footnote 84).
C.1 (C.2) causes larger incremental benefits relative to B.1 (B.2) solely through improved EU-level
comparability across Member States, stemming from the harmonised population base
(measure 2.3). Notably, large majorities of OPC respondents in all stakeholder groups (from 73%,
or 8 out of 11 respondents among statistics producers, up to 93% or 39 out of 42 respondents
among professional users) put a high priority on introducing a strictly harmonised population
definition (either for all datasets or with justified exceptions)85
. In the targeted survey with NSIs,
maintaining the status quo of non-harmonised population definitions was considered to have minor
or moderate negative impacts by a majority of the 29 respondents (69% on usability of European
data for policymaking, 79% each on comparability of national data and on quality of European
data), while production costs was the only area where a majority (55%) of respondents saw a
positive impact (38% ‘major’, 17% ‘moderate’ or ‘minor’)86
. During the option validation
84
See overall assessment of this option in the external support study (footnote 71), Section 4.5.4.
85
Responses to OPC survey (Q3.2).
86
Responses to NSI survey (Q3.3).
Figure 10 – Relative preference for policy options by stakeholder group, based on the prioritisation of policy measures expressed
in the OPC (responses to Q3.1 – Q3.17). Note that this figure only covers OPC uptake, so findings from other consultation
activities (e.g. option validation workshops mentioned in Section 6.2) complement this picture. In particular, ‘institutional users’
do not cover EU policy users, who did not submit replies to the OPC; their views were captured in a targeted survey and a
dedicated option validation workshop with the Commission. (Source: Eurostat analysis)
0,00
0,05
0,10
0,15
0,20
0,25
0,30
A B.1 B.2 C.1 C.2 D.1 D.2
Institutional users (N=9) Researchers (N=10)
Other professional users (N=28) Statistics producers (N=10)
46
workshops, most participants from NSIs as well as from among professional users found option C.1
feasible to implement, whereas the views on C.2 were split: not realistic according to a majority of
NSI participants, realistic according to professional users (see footnote 84).
D.1 brings larger incremental benefits for users than C.2, with even better EU-level comparability
through full harmonisation (measure 2.2), and a major improvement of data evidence (additional
measures 5.2, 7.2, 9.3, 14.3, 16.3). Statistics producers and Eurostat both benefit from further
simplified processes due to effective adaptation mechanisms to evolving data needs (measure 15.3)
and from full EU-level interoperability of statistical population registers (measure 17.5). In the
option validation workshops, participants from NSIs unanimously found D.1 unrealistic, driven
mostly by expectations of excessive additional burden on NSIs, while participants among
professional users generally stated that the option is realistic (see footnote 84).
D.2 causes larger incremental benefits than D.1 solely through the introduction of statistical
population registers across all Member States (measure 1.2). This benefits users with higher data
quality in general and quicker adaptation to their evolving needs. Statistics producers also benefit
substantially over time thanks to the integrated statistical production processes. In the option
validation workshops, NSI participants unanimously agreed that D.2 is unrealistic, largely due to
the costs involved in transitioning to statistical population registers and due to the subsidiarity and
proportionality concerns around them, while this option was largely viewed positively by
professional users (see footnote 84). In the OPC, a large majority (84%, or 27 out of 32
respondents) among professional user groups also prioritised the introduction of statistical
population registers, while the opinions of statistics producers were split (50%, or 5 out of 10
respondents prioritised maintaining the status quo)87
.
87
Responses to OPC survey (Q3.1).
47
Table 6 – Beneficiaries and qualitative level of incremental benefits expected by policy options compared to the baseline (scale: no increment, ‘+’, ‘++’ or ‘+++’).
Beneficiary Incremental benefits compared to the baseline B.1 B.2 C.1 C.2 D.1 D.2
People (the
public and
migrants)
Benefits from improved policymaking, including for potentially disadvantaged population groups + ++ + ++ +++ +++
Better information on own local/regional environment + ++ + ++ +++ +++
Reduced response burden, i.e. fewer direct inputs needed from individual persons +++
EU institutional
users
Better EU-level timeliness and completeness of statistics across all Member States + ++ + ++ +++ +++
Better EU-level comparability and coherence of statistics across all Member States + + ++ ++ +++
More accurate and comparable total population counts for Council voting ++ ++ ++ +++
Improved inputs to monitoring of demographic change and projecting long-term budget sustainability in
relation to population ageing
+ ++ + ++ +++ +++
Better data evidence for regional and cohesion policies ++ +++ ++ +++ +++ +++
Better data evidence for local and urban/rural policies ++ ++ +++ +++
Better data evidence for free movement of persons in the EU internal market ++ +++ ++ +++ +++ +++
Better data evidence for non-EU migration and migrant integration policies + ++ + ++ +++ +++
Better data evidence for fundamental rights and non-discrimination policies ++ ++ +++ +++
Better data evidence for the European Green Deal and housing policies +++ +++ +++ +++
Better data evidence for access to services analysis and monitoring +++ +++ +++ +++
Better data evidence for disaster/crisis response ++ ++ +++ +++
Reputational gains from improved policymaking and EU decision-making in general + ++ + ++ +++ +++
Other
institutional
users (national
and subnational
levels)
Better comparability with other Member States on population size at all levels + + ++ ++ +++
Better comparability with other Member States on migration patterns + ++ + ++ +++ +++
Better comparability with other Member States on migrant integration + ++ + ++ +++ +++
Better comparability with other Member States on housing +++ +++ +++ +++
Better comparability with other Member States on equality and non-discrimination ++ ++ +++ +++
Better comparability with other EU regions ++ +++ ++ +++ +++ +++
Better comparability with other EU municipalities and functional areas ++ ++ +++ +++
Better data evidence (through grids) for policymaking in border regions and local crisis response +++ +++ +++ +++
Reputational gains from improved visibility and transparency in a European context + ++ + ++ +++ +++
Other Better comparability of research/analysis across all Member States + + ++ ++ +++
48
professional
users
Better research/analysis of detailed geographic patterns across the EU + ++ + ++ +++ +++
Better research/analysis of migration patterns and migrant integration across the EU + ++ + ++ +++ +++
Better research/analysis of equality and non-discrimination across the EU ++ ++ +++ +++
Better research/analysis of longitudinal patterns across the EU + + +++ +++
Reduced administrative burden (through ability to find all statistics on Eurostat’s website) + ++ + ++ +++ +++
Economic benefits from availability of better European statistics + ++ + ++ +++ +++
Benefits from improved policymaking (e.g. enhanced social cohesion) + ++ + ++ +++ +++
Statistics
producers
(NSIs)
Improved comparability of statistics with other Member States + + ++ ++ +++
Improved accuracy and coverage of statistics due to interoperability with other Member States ++ +++
Reduced administrative burden (through simplified statistics transmission processes) + + + + + +++
Reduced administrative burden (through integrated statistics production process) +++
Reduced administrative burden (through improved use of administrative and/or other data sources) ++ ++ ++ ++ ++ +++
Reduced administrative burden (relating to regulatory changes to adapt to evolving policy needs) ++ ++ +++ +++
Increased ability to meet legal requirements + + + + + +++
Increased staff skills + ++ + ++ ++ +++
Administrative
data providers
Improvements in administrative registers thanks to closer collaboration with NSIs ++ ++ ++ ++ ++ +++
Reduced administrative burden through streamlined data exchange with NSIs ++ ++ ++ ++ ++ ++
Increased added value from own data through improved reuse ++ ++ ++ ++ ++ +++
Improved legal base of statistical cooperation through a clear mandate +++ +++ +++ +++ +++ +++
Reputational gains from improved reuse of administrative registers ++ ++ ++ ++ ++ +++
Eurostat Advancement of Eurostat mission ‘to provide high-quality statistics and data on Europe’ + ++ + ++ ++ +++
Improved collaboration with EU policy users + ++ + ++ +++ +++
Improved collaboration with ESS partners due to EU-wide interoperability ++ +++
Better data evidence for other European statistics (e.g. data collection based on samples, national
accounts)
+ + ++ ++ ++ +++
Reduced administrative burden (relating to regulatory changes to adapt to evolving policy needs) ++ ++ +++ +++
Reduced administrative burden (related to coordination / quality assurance for voluntary data) ++ ++ +++ +++
Reputational gains from improved international standing of European statistics freely available to all + ++ + ++ +++ +++
49
Table 7 – Comparative overview policy option assessments (see Annex 2 part 1.1 for detailed quantitative cost estimates and part 1.2 for detailed qualitative assessment scores)
Dimension Policy options
A B.1 B.2 C.1 C.2 D.1 D.2
Proportionality Pass Pass Pass Pass Pass Uncertain Uncertain
 Necessary minimum Pass Pass Pass Pass Pass Pass Pass
 Limitation to EU scope Pass Pass Pass Pass Pass Pass Uncertain
 Costs commensurate Pass Pass Pass Pass Pass Pass Pass
 Simplest effective action Pass Pass Pass Pass Pass Uncertain Uncertain
Effectiveness – average 0.00 0.93 1.93 1.21 2.21 3.14 3.71
 Specific objective 1 0 1 1.5 2 2.5 3 4
 Specific objective 2 0 1 2 1 2 3 4
 Specific objective 3 0 1 3 1 3 3.5 4
 Specific objective 4 0 1 2 1 2 3.5 4
 fundamental rights 0 0.5 1 1.5 2 3 3
 society 0 1 2 1 2 3 3
 environment 0 1 2 1 2 3 4
Coherence with EU objectives 0 1 2 2 3 4 4
Efficiency Baseline Ambition in incremental effectiveness grows with incremental costs Better than D.188
 total incremental costs89 0 10.5 – 15.7 42.0 – 63.0 31.5 – 47.2 63.0 – 94.5 95.0 – 142.5 145.6 – 218.490
Views of key stakeholder groups Big difference of views between statistics producer and user profiles
 Statistics producers Feasible Split views Feasible Split views Not feasible (unanimous)
 EU and other institutional users Feasible but outputs not sufficient for current policy needs Barely sufficient Feasible and sufficient
 Researchers, other professional users Feasible but outputs not sufficient for professional needs Split views Feasible and sufficient
88
According to the detailed comparative efficiency assessment in Annex 2 part 1.2 (see Subsections 1.2.6 for D.1 and 1.2.7 for D.2), policy option D.2 is expected to be more efficient than D.1
in the long term due to the gradual cost savings and subsequent efficiency gains incurred through statistical population registers over time (this may stretch over a long period going beyond the
10 years assessed here). While D.1 is expected to be less efficient than the more modest options (including notably C.2) due to the extremely ambitious increase in statistical outputs without
taking strong facilitating action, the relative efficiency of D.2 against the more modest options remains unclear (see further description in Section 7).
89
+/- 20% margin in millions of 2021 EUR. As shown in Table 3, most of these estimated incremental costs (97.7% to 98.5% depending on the policy option) fall on Member States, mostly
in the form of additional human and financial resources required by NSIs and the respective national statistical production systems.
90
Specifically for policy option D.2, almost all incremental costs with respect to D.1 are incurred by Member States not currently having or planning a statistical population register (group 3 in
Section 6.1.1). Potential cost savings estimated in Section 6.1.3 have not been set off against the costs here, due to larger uncertainties around these estimates.
50
7. 7. HOW DO THE OPTIONS COMPARE?
In line with the Better Regulation guidelines (BRG)91
, a multi-criteria decision analysis (MCDA)
should be used to compare the different options in terms of compliance with the proportionality
principle, effectiveness against specific objectives and wider impacts (fundamental rights, social,
environmental), coherence with EU objectives, and efficiency. Table 7 summarises the assessment
of all policy options for each of these criteria, by also including views of key stakeholder groups. A
detailed reasoning for each of the assessments, presented by policy option, is provided in Annex 2.
A comparison of policy options synthesising these findings is presented in this section.
On proportionality, four questions were answered (see also Section 4.6.1 of the external support
study, footnote 71):
1. Does the option go beyond what is necessary to satisfactorily achieve the objectives?
All options pass, as all policy measures considered are clearly linked (albeit with varying
ambitions) to specific problems with the current EU legal framework, as established by the
evaluation.
2. Is the scope of the option limited to those aspects that Member States cannot achieve
satisfactorily on their own and where the EU can do better?
Member States alone clearly cannot provide European statistics fulfilling required common
quality standards, especially on completeness, comparability, frequency, timeliness and
geographic and topical detail, and thus relevance for EU policy needs.
Policy options B.1 through D.1 pass, as all their measures are clearly linked to creating or
improving such common standards. The status of option D.2 is uncertain due to measure 1.2
requiring statistical population registers in all Member States.
While the goals of improving efficiency, flexibility and comparability across all Member
States are proportionate, some NSI experts have expressed concerns that national systems
would lose flexibility to find their own solutions92
. However, it is the freedom of movement
in the EU internal market under EU law that presents characteristic challenges for statistics,
such as highly dynamic and further increasing mobility, cross-border work/family patterns
and several places of residence split across regions and countries becoming common
practice.
Experience over the past years suggests that national approaches are ineffective to produce
genuine European statistics93
, so that EU action would be proportionate in the form of
(ideally interoperable) national population registers solely for statistical purposes.
3. Are costs for the EU, national governments, regional or local authorities, economic
operators or the public commensurate with the objectives of the initiative?
Based on the quantitative analysis of the costs (Section 6.1), all policy options are judged as
passing this question. This includes the costliest option, policy option D.2, because a large
part of these costs (around the statistical population registers) would be an investment in the
91
European Commission (2021) Better Regulation Toolbox Chapter 8, Tool #62.
92
At the option validation workshop with NSIs and on various other regular expert consultations in the ESS.
93
For instance, after the adoption of the Demography Regulation, Eurostat initiated studies with NSIs on how the
population definition based on usual residence can be implemented within the national systems. The outcome is that
currently there is no common implementation of this definition, which leads to the continuing situation that the EU
population cannot be measured at a single place of residence (or main residence) across the entire EU territory, and
there is a constant risk of under or double counting people, vital events and migration asymmetries. In absence of a
single European identifier, addressing such issues is generally difficult for NSIs, but there are methods based on data
sharing between NSIs that are worth exploring. In the longer term, additional benefits could emerge from improved
European digital identities under a proposed amendment to Regulation (EU) No 910/2014; COM(2021) 281.
51
capacities of a few Member States (group 3 in Section 6.1) that is expected to be offset by
long-term savings and wider benefits at EU level.
4. Is the form of action (choice of instrument) as simple as possible and coherent with the
satisfactory achievement of the objective and effective enforcement?
The choice of legal instrument (one new regulation or an update of existing legislation) to
implement the policy options has not been explored in depth in this study. Any of the
measures entailed can, in principle, be achieved by each approach, whereas integrating the
current three basic acts into one would have natural benefits for streamlining all related
processes. Thus options B.1 through C.2 pass this question, while whether D.1 and D.2 pass
remains uncertain: D.1 is so ambitious on statistical outputs that there are increased risks of
significant non-compliance across Member States. The statistical population registers
required under D.2 can be seen as facilitating the statistical infrastructure in this respect. It
is known from experience that statistical population registers are not simple to implement
and may compete with other measures that address comparability issues at EU level (e.g.
harmonised population definition, internal migration flows)94
.
Regarding effectiveness in achieving the specific objectives (SO), all options were scored against
the relevant policy measures included (see Figure 9) for each SO.
 SO1 (completeness, coherence, comparability) took into account policy measures to
harmonise the population definition (policy topic 2), ensure completeness of voluntary
statistics (topics 12, 13) and ensure a consistent time series (topic 16). Moreover, statistical
population registers (topic 1) were considered to reduce implementation risks of a
harmonised population definition. Therefore, option D.2 scored highest (+4), followed by
D.1 (+3) and C.2 (+2.5).
 SO2 (timeliness and frequency) took into account measures to improve the publication
frequencies (topics 5, 7) and the time between reference and publication dates (topics 8, 9)
of various statistical products. Statistical population registers (topic 1) were considered to
improve production efficiency and thus reduce compliance risks. Option D.2 scored highest
(+4), followed by D.1 (+3) and by B.2 and C.2 (+2).
 SO3 (geographic and topical detail) took into account measures to improve the geographic
detail of existing statistics (topics 3, 4, 6) and introduce new statistics or more details on
topics of key policy interest (topics 10, 11, 14). Statistical population registers (topic 1)
would allow georeferencing of the whole population and also improve the availability of
statistical detail and thus reduce compliance risks. Option D.2 scored highest (+4), followed
by D.1 (+3.5) and B.2 resp. C.2 (+3).
 SO4 (input and output flexibility) took into account measures to adapt the statistical
products more quickly and efficiently to evolving policy needs (topic 15) and enable better
use of existing or new data sources that may become available in the future (topic 17).
Statistical population registers (topic 1) would streamline the integration of data sources and
also be more flexible to extract new statistical outputs, and thus reduce compliance risks.
Option D.2 scored highest (+4), followed by D.1 (+3.5) and B.2 resp. C.2 (+2).
Options were also scored against wider impacts on fundamental rights, society and environment.
While the scores generally followed the patterns of SO scores, the impact of statistical population
94
For instance, instead of setting up statistical population registers, a thorough scaled-up auditing of national practices
by Eurostat, including regular country visits to extensively assess national production systems, could be considered – at
least in principle – to give stronger incentives for common operationalisation and implementation of the harmonised
concepts.
52
registers in option D.2 were particular, in that some concerns were expressed about certain risks of
negative impacts on fundamental rights (related to personal data protection) and social impacts
(doubts around the social licence). However, they would have specific positive environmental
impacts by embracing digitalisation and efficient production systems, thus reducing the carbon
footprint of statistics production, especially when full field enumerations for traditional censuses
become obsolete. Overall, on effectiveness, option D.2 scored highest (+3.71) followed by D.1
(+3.14) and C.2 (+2.21).
Regarding coherence with EU objectives, all options were scored against the relevant policy
measures on how ambitiously statistics move towards EU-level comparability of population
definitions (topic 2), deliver on the data needs for current EU priorities (topics 3, 6 as regards
urban/rural integration and the Green Deal; topics 4, 10, 11 for migrant integration; topic 14 for
equality) and improve flexibility to adapt to evolving policy needs and data environments (topics 1,
15, 17 for enabling new data sources and interoperability of ESS production systems). Overall, on
coherence options D.1 and D.2 scored the highest (+4) followed by C.2 (+3) and B.2 resp. C.1 (+2).
Regarding efficiency, baseline and incremental costs by policy option have been quantified in
Section 6.1. The baseline (A) sets zero incremental costs or benefits but is considered dynamic in
that key impacts of modernisation already planned in some Member States (group 2 of Section 6.1)
are factored in. However, most of the benefits could not be quantified, so incremental costs and
benefits by option can only be confronted qualitatively. Nevertheless, three broad qualitative trends
can be identified.
1. Incremental effectiveness grows steadily with incremental costs. This means that a simple
retail analogy applies: the more expensive option buys more or bigger benefits – in
increasing order, this is B.1, C.1, B.2, C.2, D.1 and D.2. However, a key aspect to note is
that most costs are incurred by statistics producers, while the most relevant direct
beneficiaries are statistics users (see Table 6). Nonetheless, at a higher level, both costs and
direct benefits are incurred by the same actors, namely public administrations across the
EU. Researchers and other professional users also benefit directly from better statistics
availability, while society as a whole benefits from key indirect benefits, mostly through
better policymaking.
2. Introducing a harmonised population definition based on a strict usual residence concept
(going from option B.1 to C.1 or from B.2 to C.2) is estimated to be almost as expensive as
all the significant improvements on statistical output and flexibility together (going from
B.1 to B.2 or from C.1 to C.2). This is because, even in advanced integrated statistical
production systems based on administrative registers, complex specific efforts are needed to
capture the actual presence of persons at a registered place of residence (i.e. control over-
coverage) and the size and characteristics of population groups actually present on a given
territory but not registered there (i.e. control under-coverage). Nonetheless, improving on
this aspect has a leverage effect on higher quality, not only of population outputs from this
initiative, but of all European social statistics (since accurate enumeration of the whole
population is the backbone of other statistical tools such as collecting data from samples).
3. The respective high incremental costs and burdens during an extended adjustment phase of
options D.1 and D.2 along the lines of points 1 and 2 above are clear, but despite associated
large benefits, the extent to which these outweigh the costs is uncertain. Nonetheless, option
D.2 is considered more efficient than D.1, as they both have the same statistical outputs and
flexibility, but statistical production in D.2 is expected to be more efficient in the long term
due to the full process integration through statistical population registers. However, the
monetised cost and benefit estimates for option D.2 are very uncertain, meaning it is unclear
over which period the net recurrent benefits break even with the required one-off
53
investments. In reality, D.2 is expected to be much more efficient than the baseline, maybe
even than all options B-C (especially as recurrent cost savings accrue), but a conservative
assessment requires particular caution on this option, which may be very costly upfront for
some Member States.
The views of key stakeholder groups on the different policy options mainly reflect the insights on
effectiveness v costs (see point 1 of the efficiency assessment above). Most statistics producers
consulted during the various consultation activities (OPC, targeted survey, two dedicated
workshops) focused exclusively on the additional costs and burden on their own organisation.
However, most professional statistics users consulted (institutional users, researchers and other
professional users) put emphasis primarily on the benefits of upgraded statistical outputs and
flexibility. This led to a deep division between these stakeholder profiles. The producers strongly
favoured the status quo (baseline) or, at most, options B.1 or C.1, while the users strongly favoured
options D.1 or D.2 or, at least, C.2. This split between producers and users is clearly visible in
Figure 10 (illustrating OPC uptake) and in Table 7 (integrating the uptake from all consultation
activities)95
.
8. 8. PREFERRED OPTION
8.1. 8.1. Overall ranking and selection of the preferred option
Table 7 summarises the comparative performance of all policy options against the criteria assessed.
With a lack of quantified benefits, a direct ranking of options is not possible. The efficiency
assessment, however, showed qualitatively that none of the options is obviously more cost-effective
than any other. Rather, the options offer increasing benefits (directly for statistics users and
indirectly for the whole society) at increasing costs (mostly for statistics producers, i.e. national
statistical production systems). The deep division between producers and users of statistics reflects
this pattern, as producers focused on costs while users prioritised the benefits. Option C.2 appears
as a possible compromise because costs are still considered high by producers (but not by too
much), while benefits are considered barely sufficient by users (but still an improvement compared
to the current situation).
The underlying policy decision to be made is which level of ambition – and costs entailed – is
deemed appropriate for a more future-proof EU framework for population statistics. The main
concerns of statistics producers need to be read in this context as driven mainly by a strict limitation
of new costs and burden. However, the assessment has clearly shown that ambitious action on data
needs for EU policy priorities has its price, in the form of additional resources needed for statistics
producers that are substantial compared to the current baseline costs (up to around 10% for
option D.2 according to Table 3). More precisely, only the most ambitious options, D.1 and D.2,
contain strong measures to address the needs of key EU policy areas like urban/rural integration,
the Green Deal, and fundamental rights and non-discrimination. Additionally, only option D.2
includes statistical population registers as a strong measure to increase production efficiency and
thus facilitate delivering the ambitious output goals.
95
Unfortunately, the dedicated option validation workshops (see external support study, footnote 71) failed to reconcile
these perspectives. In particular, the workshop with NSI experts did not distinguish clearly incremental costs and
burden from technical and methodological feasibility aspects.
54
Therefore, the overall preferred option is D.2. While option D.2 is the most ambitious in terms of
statistical output and flexibility of the framework, it achieves the best result thanks to a similarly
ambitious simplification and integration of the statistical production systems and sustainable long-
term efficiency gains. However, the main issues of this option argued in Section 7 are the
uncertainties around subsidiarity and proportionality, and the significant incremental costs and
benefits of introducing interoperable statistical population registers in all Member States.
Therefore, an alternative (conservative) approach preferring option C.2 would also be
reasonable if the acknowledged sizeable uncertainties around the proportionality and efficiency of
option D.2 are given more weight – this would also be more acceptable to statistics producers as
key stakeholders for implementation.
Finally, as pointed out in Section 4.2, each option was constructed bottom-up from a combination
of more specific policy measures addressing the specific objectives. This means that all options
represent a form of necessary simplification among the available measure combinations. While
option D.2 represents a group of measures that collectively emerged as the preferred package
among all assessed, when further developing this intervention it may be reasonable, proportionate
and ultimately necessary to consider adjustments around this package at the level of individual
measures. However, the unique feature of option D.2 – and a significant facilitating factor for its
very ambitious output goals – is the introduction of interoperable statistical population registers in
all Member States, which makes this option the most ambitious one, not only in terms of
requirements and outputs, but also in terms of expected long-term simplification, integration and
efficiency gains. Therefore, a legislative proposal in line with option D.2 should include credible
ambition in this direction.
8.2. 8.2. REFIT (simplification and improved efficiency)
All options going beyond the baseline, including the preferred options C.2 or D.2, would entail an
update of the current legal framework and thus make some level of streamlining the legal base
relative to the current situation possible. This may give rise to benefits in terms of compliance,
monitoring and enforcement, and associated cost savings.
For instance, it is expected that some cost savings could be made by reducing the current
administrative burdens borne by the Commission (Eurostat) as well as Member States in all groups
and their respective NSIs. It was not possible within the scope of this study to quantify this effect
due to uncertainties around the overall cumulative effect that streamlining the legal framework
would have on individual Member States. For example, it is possible that the individual
administrative approach taken by each Member State to provide and report on population statistics
is embedded within other procedures, tempering the cost-saving effect from streamlining the legal
base.
More generally, given the nature of the measures proposed (i.e. the fact that most measures required
placing additional requirements and therefore burdens on statistics producers and thus on public
administrations), net simplifications and associated savings were challenging to quantify.
Nonetheless, Table 8 summarises the qualitative analysis of the relevant benefits under the
preferred option.
Notable simplifications are incurred by introducing flexibility features leading to simplified data
sharing between source data owners and NSIs as well as to simplified regulatory adaptations of
NSIs and Eurostat to evolving data needs. Users will benefit from simplified and centralised access
to all relevant data through the Eurostat website. Moreover, both NSIs and Eurostat will also
benefit from simplified data transmission procedures from streamlined legislation.
55
Finally, the introduction of statistical population registers in all Member States (measure 1.2,
uniquely appearing in option D.2) is expected to bring about significant long-term efficiency gains
due to a substantial simplification and full integration of national statistical production processes.
The full volume of potential cost savings is very hard to quantify, and will depend on the specific
national backgrounds of the Member States that currently do not have such a register and are not
planning one (group 3 in Section 6.1). Nevertheless, this impact assessment has set a benchmark by
roughly estimating the potential recurrent EU-level cost savings on censuses to between
EUR 141 million and EUR 563 million per census round (Section 6.1.3).
In practice, Member States in group 3 that do not currently intend to set up statistical population
registers might also accrue recurring cost savings in relation to the delivery of other types of data
requirements, including annual statistics. However, these impacts could not be reliably quantified
since no data were available on the isolated effect of using statistical population registers for annual
statistics specifically. To this end, the estimate for cost savings associated with the census rounds is
a proxy for the wider cost savings associated with the use of statistical population registers. Since
the cost savings related to the census was significant for those Member States that do not currently
have a population register, estimating this value was deemed to be the most proportionate approach.
8.3. 8.3. Application of the ‘one in, one out’ (OI-OO) approach
8.3.1. 8.3.1. Potential new burden on citizens
Regarding any new burden on citizens, the only potential source of such impacts under the
preferred options D.2 or C.2 is policy measure 14.3 (only in D.2), introducing new collection
modes for equality data. Since at least some of these variables would need to be collected by self-
declaration (see footnote 49), some form of incremental direct interaction with citizens (e.g. surveys
conducted in person, by phone or online) will generally be required. However, due to the relative
stability of respective population group patterns, such data are expected to be needed only in less
frequent intervals, for instance during census years. For instance, according to the targeted NSI
survey, in the 2021 census round 20 Member States relied on either full enumeration (13 countries)
or sample surveys (7 countries) as part of their census taking modes96
. Moreover, according to a
Eurostat monitoring survey (state September 2021), in total 19 countries will publish data on at
96
Responses to NSI survey (Q2.1); Greece imputed from https://www.statistics.gr/en/2021-census-pop-hous.
Table 8 – Summary of REFIT relevant benefits of the preferred option D.2 and alternative C.2.
Beneficiary Reduced administrative burden through: D.2 C.2
Professional users ability to find all needed statistics on Eurostat website +++ ++
Statistics producers
(NSIs)
simplified statistics transmission processes +++ +
integrated statistics production process +++
improved use of administrative and/or other data sources +++ ++
regulatory changes to adapt to evolving policy needs +++ +++
Admin. data providers streamlined data exchange with NSIs ++ ++
Eurostat regulatory changes to adapt to evolving policy needs +++ ++
coordination/quality assurance for voluntary data +++ ++
56
least one of the equality variables other than sex and age with the 2021 census results, four of them
based on a fully register-based census97
.
The incremental burden on citizens will be mainly measured by additional time afforded by survey
respondents to answer the relevant new questions. Now the point is that sample surveys are a
standard instrument of data collection carried out regularly also at EU level (e.g. annually for EU-
SILC or even quarterly for the EU-LFS), with considerable baseline burden on citizens in terms of
average person-time spent per year. For example, according to the impact assessment on Regulation
(EU) 2019/170098
, a typical LFS interview covering roughly 100 questions takes about 20 minutes
(median). At a typical sample size of around 1% of the population per quarter99
, this means a total
response burden of very roughly 6 million person-hours per year EU-wide (order of magnitude).
Going further, the following assumptions can be made for collecting equality-relevant population
data other than sex and age based on a sample survey:
 a roughly similar setting as EU-LFS is needed in terms of sample size per survey (1%) and
question duration (0.2 minutes per question)100
;
 an equality survey would be conducted (or respective module added to an existing survey)
multiannually, e.g. once every 10 years during census years;
 about 10 extra questions would be needed (two questions per five equality dimensions other
than age).
As a result, the total incremental time burden on citizens would be a fraction of 1/400 or 0.25% of
the regular baseline burden caused by EU-LFS only.
However, there are issues specific to surveys or survey questions addressing grounds for
discrimination, namely that the overall response rate drops with each question potentially perceived
as contentious or sensitive by a sizeable share of respondents101
. This is a known challenge
normally requiring larger sample sizes and thus overall response time to achieve the same level of
accuracy (e.g. compared to EU-LFS). Therefore, one could allow a very conservative uncertainty
margin on the result above, for instance estimating the relative incremental burden as up to 1% of
the current baseline burden. It can still be concluded that a new or extended direct population data
collection at EU level – under the assumptions reflected above – would cause only negligible
incremental response burden on citizens and would thus not generate any significant net ins
relevant for OI-OO.
97
This means these countries have some of the equality variables in national administrative sources, e.g. non-binary
gender (AT), ethnicity (LT, LV), religious affiliation (FI, LT).
98
SWD(2016) 283.
99
https://ec.europa.eu/eurostat/statistics-explained/index.php?title=Labour_force_survey_(LFS)_-
_sampling_design,_sample_size_and_sampling_errors#What_is_the_sample_size_in_each_country_.3F
100
Notably, the target groups of EU-LFS and such a new survey would not be fully the same, as EU-LFS addresses
private households only, whereas target groups at risk of discrimination or inequality do not necessarily live in regular
private households. However, this intricacy does not affect the ensuing argument on rough total response time in the
whole population. (It does affect the potential incremental costs on NSIs carrying out such a survey, but this is not
relevant here.)
101
In a Eurobarometer survey, respondents expressed support for providing sensitive personal details on an anonymous
basis, if that could help to combat discrimination in their country. This applies to information on their ethnic origin
(72% in favour), their religion or beliefs (71%), their health situation (66%) and their sexual orientation (63%), This
shows that there are sizeable minorities who still have reservations against such questions, which supports the
assumptions that the non-response would typically increase. Special Eurobarometer on Discrimination in the EU in
2015: https://op.europa.eu/en/publication-detail/-/publication/d629b6d1-6d05-11e5-9317-01aa75ed71a1.
57
8.3.2. 8.3.2. Potential new costs for businesses
Regarding new costs for businesses, the only potential source of such impacts under the preferred
options D.2 or C.2 is policy measure 17.3, enabling B2G data-sharing approaches to European
population statistics. In this context, the impact assessment on the Commission’s initiative on a
Data Act102
has assessed relevant costs and benefits of such approaches to official statistics in
general. A summary of these is presented in Table 9 and Table 10.
The direct benefit in Table 9 relates to the fact that statistical authorities in the ESS try any way
they can to maximise the use of new sources from private holders of data (PHD). Such factual
requests, partly based on national laws, already create a factual procedural burden, and thus costs,
that are driven further by redundancies and inefficiencies in the absence of a legal base. Clear
common enablers across the EU as included in the Data Act proposal are thus estimated to lead to a
102
SWD(2022) 34.
Table 9 – Overview of B2G direct benefits related to the Data Act proposal (impact assessment (IA) preferred policy option) – from
Table I in Annex 3 to the IA (footnote 102)
Overview of B2G incremental benefits related to the Data Act proposal (IA preferred option)
Description Amount Comments
Direct benefits
Lower administrative
burden
EUR 155 million per year Large and medium businesses should
experience lower compliance costs
and less duplication in B2G data
sharing.
Qualitative benefits include improved
reputation and workforce motivation.
Table 10 – Overview of B2G incremental costs related to the Data Act proposal (IA preferred option) – from Table II in Annex 3 to
the IA (footnote 102)
Overview of B2G incremental costs related to the Data Act proposal (IA preferred option)
Citizens/Consumers Businesses Administrations
One-off Recurrent One-off Recurrent One-off Recurrent
B2G data
sharing
Direct costs n/a n/a EUR 552.5 m EUR 78.1 m n/a EUR 21.6 m
Indirect costs n/a n/a n/a n/a n/a n/a
58
direct saving of administrative costs of EUR 155 million per year on the PHDs to be shared under
B2G. However, the direct costs on PHD businesses in Table 10 consist of (one-off) adjustment
costs of EUR 552.5 million and (recurrent) administrative costs of EUR 78.1 million per year.
A cost-benefit analysis focusing on the OI-OO relevant impacts incurred by the PHD sector
suggests that the benefit mentioned above affects the PHD sector directly and exclusively, which
allows for direct compensation and offsetting of OI-OO relevant costs. In particular:
 administrative costs (recurrent) are directly offset by administrative cost savings
introduced by the same initiative, leaving recurrent excess cost savings of EUR 76.9 million
per year;
 adjustment costs (one-off) are compensated over time by the recurrent offsetting excess
cost savings above; more specifically, just over 7 years would be needed, based on the
estimated amounts, for the entire PHD sector to fully compensate their adjustments.
According to this analysis, B2G data sharing in general, including for official statistics, is not
expected to generate OI-OO relevant net ins on businesses. This argument can be projected to
specific sectors of official statistics – such as population statistics – by assuming a linear scaling of
both benefits and costs from B2G by the same factor. In this model, it is not even necessary to
estimate the specific scaling factor for population statistics because the OI-OO
offsetting/compensation relation above is also linear and thus not sensitive to the value of the
scaling factor.
9. 9. HOW WILL ACTUAL IMPACTS BE MONITORED AND EVALUATED?
The performance of the new ESOP legislation will be monitored and evaluated against the
operational objectives (OOs) set out in Table 11. The following paragraphs outline the main
considerations for monitoring and evaluating each stage of the policy cycle. These will complement
the regular monitoring and quality control mechanisms already present in the ESS that are mostly
based on regular quality reports by NSIs. It has become a default standard to anchor such quality
reporting in EU statistical legislation, and this initiative aims to further strengthen and specify the
legal basis on quality reporting and monitoring. Additionally, surveys with key user groups have
shown to be useful in the stakeholder consultation for this initiative. These could become regular
parts of the monitoring approach to track the user uptake of the updated statistical outputs.
During the implementation phase of the new legal framework, the Commission (Eurostat) will
continue organising regular expert group meetings with partner NSIs in the ESS to discuss and
clarify any issues that may arise, continuing a long-standing history of good and close cooperation
between Eurostat and its ESS partners on technical and statistical matters. This includes diligent
joint preparation of key implementing acts regulating the detailed new statistical data and metadata
requirements, which will be of central interest to both statistics users and producers. For instance,
Eurostat can establish topical task forces to address specific technical challenges in achieving the
objectives of the new legislation. The implementation stage is planned to conclude with a first
evaluation focusing on the implementation, functioning and initial impacts of the new legal
framework. To obtain sufficient information on the performance, this evaluation is planned within 3
to 5 years after the entry into force of the new legal framework, in line with BRG advice that
evaluations should have access to at least 3 full years of data.
After transition to the application phase, the Commission (Eurostat) plans to evaluate the
functioning and impact of the legislation every 3 to 5 years. Table 11 below outlines some potential
59
indicators to monitor the effectiveness of a new ESOP framework against each OO, including
potential data sources that that are easily available and credible. In this context, the recent
evaluation of the current legislation (see footnote 7) has developed several highly valuable
indicators – in particular regarding statistical quality based on data and metadata published by
Eurostat – that will be considered for consecutive monitoring of the new legal status quo. In
addition, Eurostat’s regular user satisfaction survey, along with targeted surveys (e.g. of NSIs and
statistical correspondents of Commission services), have proven to be good tools for evidence-
gathering during the past stakeholder consultation, and are thus considered valuable instruments for
future evaluations.
60
Table 11 – Possible key performance indicators, targets and data sources for the operational objectives (OO). Indicators reused from the recent evaluation of the current legislation are referenced by their
acronyms used in Annex 5 of the evaluation report (e.g. ‘SQ1.4’ for ‘statistical quality indicator number 1.4’ and so on – see evaluation report). Current benchmarks are taken from the evaluation report
(footnote 7) and from stakeholder consultation findings, otherwise the source is referenced explicitly.
SO OO Possible key performance indicators and data sources Current benchmarks Targets
1 Complete,
coherent and
comparable statistics
1.1 Coherent processes
used to produce
statistics across
Member States (MS)
to improve
comparability
KPI 1.1.1 Number of MS producing their data with a single
integrated process
Sources: Eurostat metadata, quality reports, targeted
surveys with NSIs
12 MS All MS after 10 years from
adoption of the new legal
framework
KPI 1.1.2 Evaluation SQ1.4: number of MS providing
coherent annual and census data for 2031
Source: Eurostat data
22 MS All MS
KPI 1.1.3 Evaluation SQ1.5: number of MS providing
consistent demographic balances
Source: Eurostat data
12 MS All MS after end of
transition period
1.2 Consistent
population base
definitions used across
MS and within
datasets submitted by a
MS
KPI 1.2.1 Evaluation SQ1.1: number of MS using the strict
usual residence definition for their population base
 Sources: Eurostat metadata, quality reports, targeted
surveys with NSIs
18 MS All MS after end of
transition period
KPI 1.2.2 Number of MS providing qualitative metadata on
how they implement the population base definition
 Sources: Eurostat metadata, quality reports
0 MS103
All MS
KPI 1.2.3 Number of MS providing quantitative metadata
on how accurately they implement the population base
definition
0 MS (footnote 103) All MS
103
Eurobase metadata accompanying the dataset on usually resident population on 1 January (demo_urespop): https://ec.europa.eu/eurostat/cache/metadata/en/demo_urespop_esms.htm
61
SO OO Possible key performance indicators and data sources Current benchmarks Targets
 Sources: Eurostat metadata, quality reports
1.3 Complete
transmission of
mandatory datasets
from all MS
KPI 1.3.1 Evaluation SQ6.1: completeness of mandatory
statistics
 Source: Eurostat data
98.9% All MS provide all
mandatory breakdown
categories after end of
transition period
KPI 1.3.2 Evaluation EE2.1: share of EU aggregates
provided in mandatory datasets
 Source: Eurostat data
84% All mandatory breakdowns
include an EU aggregate
category after end of
transition period
KPI 1.3.3 Evaluation EE2.2: share of ‘unknown’ in
mandatory breakdowns
 Source: Eurostat data
~0.1% No more than 5% of any
population broken down by
a variable where the
category ‘unknown’ exists
are assigned to that
category
1.4 Update historic
time series
KPI 1.4.1 Number of MS having updated the last 10
reference years before the 2021 census
 Source: Eurostat record of data revisions
10 MS (2011 census)104
All MS where the 2021
census requires revisions of
the last 10 reference years
KPI 1.4.2 Number of MS having updated the complete time
series
 Source: Eurostat record of data revisions
0 MS All MS after end of
transition period
2 Timely and
frequent statistics
meeting user needs
2.1 Frequencies and
deadlines set out in EU
legislation better meet
users’ needs
KPI 2.1.1 Opinion of key user groups including policy and
institutional users
 Sources: Eurostat user satisfaction survey, targeted surveys
n/a A majority of policy and
institutional users agree that
that frequency and
timeliness have improved
104
Auxiliary information assessed for indicator SQ3.4 of the evaluation report (footnote 7).
62
SO OO Possible key performance indicators and data sources Current benchmarks Targets
with Commission statistical correspondents after the implementation of
the new legal framework
KPI 2.1.2 Evaluation SQ4.1: EU data timeliness evolution
 Source: Eurostat record of data transmissions
397 days Time elapsed between
reference period and
publication of EU-complete
data becomes shorter than
baseline for all mandatory
datasets under the new
legal framework
KPI 2.1.3 Evaluation SQ4.2: EU data timeliness compared
to national and international practices
 Sources: Eurostat record of data transmissions, metadata
of national and international data collections
22 MS (population)
18 MS (births)
17 MS (deaths)
20 MS (migration)
All MS achieve timeliness
standard T+6 months
currently achieved by a
majority of MS in national
data collections
2.2 MS compliance
with legal deadlines
KPI 2.2.1 Evaluation SQ5.1: punctuality of MS data
transmissions
 Source: Eurostat record of data transmissions
31 days All MS transmit mandatory
datasets within the agreed
legal deadlines after end of
transition period (i.e. 0
days)
3 Statistics are
sufficiently
comprehensive in
terms of policy
relevant topics or
groups and
sufficiently detailed
in terms of
characteristics and
breakdowns
3.1 Statistics better
meet users’ needs
regarding:
 migrants and
migration flows
 specific
population groups
characteristics
 geographical
detail of statistics
 housing
 Opinion of key user groups including policy and
institutional users regarding:
KPI 3.1.1 Migrants and migration flows
KPI 3.1.2 Specific population groups characteristics
KPI 3.1.3 Geographical detail of statistics
KPI 3.1.4 Housing
 Sources: Eurostat user satisfaction survey, targeted surveys
with Commission statistical correspondents
n/a A majority of policy and
institutional users agree that
statistical detail has
improved after
implementation of the new
legal framework
63
SO OO Possible key performance indicators and data sources Current benchmarks Targets
4 Legal framework
is sufficiently
flexible to adapt data
to quickly evolving
policy needs and
opportunities from
new data sources
4.1 Improved
flexibility to adapt
statistical outputs to
evolving policy needs
KPI 4.1.1 Number of new datasets that were regulated
using the flexibility mechanism over time
 Sources: Executive acts adopted by the European
Commission
0 At least one successful
application of the flexibility
mechanism during the first
10 years after adoption of
the new legal framework
KPI 4.1.2 Opinion of key user groups including policy and
institutional users
 Sources: Eurostat user satisfaction survey, targeted surveys
with Commission statistical correspondents
n/a A majority of policy and
institutional users agree that
flexibility regarding new
data needs has improved
10 years after the adoption
of the new legal framework
KPI 4.1.3 Opinion of NSIs on effectiveness and
proportionality of the flexibility mechanism
 Source: targeted surveys with NSIs
n/a A majority of NSIs agrees
that flexibility regarding
new data needs is effective
and proportionate
4.2 Dynamic and
flexible use of relevant
data sources available
or becoming available
KPI 4.2.1 Number of new data sources included in
statistical production over time
 Sources: Eurostat documentation, metadata, quality and
specific reports
0 At least one new data
source is successfully
included in production in
the first 10 years after
adoption of the new legal
framework
KPI 4.2.2 Opinion of NSIs on improved availability of
relevant source data and/or cooperation with data owners
 Source: targeted surveys with NSIs
n/a A majority of NSIs confirm
improved availability of
relevant source data and/or
cooperation with data
owners
64
ANNEX 1: PROCEDURAL INFORMATION
10. 1. LEAD DG, DECIDE PLANNING/CWP REFERENCES
Lead DG Eurostat
Decide Planning PLAN/2021/10584105
CWP reference CWP 2022 Annex II - REFIT
11. 2. ORGANISATION AND TIMING
After political validation of the ESOP initiative in February 2021, an interservice steering group
(ISG) chaired by Eurostat and composed of representatives of 16 Commission DGs106
was set up to
supervise the progress on combined evaluation and impact assessment including stakeholder
consultations. The ISG met six times:
Meeting date Topics discussed
31/03/2021  Introduction to European population statistics
 Draft evaluation roadmap / inception impact assessment (IIA)
 Draft consultation strategy
 Draft terms of reference for a tender on evaluation/IA support
20/08/2021  Introduction of contractor ICF SA for support study
 Progress on evaluation/IA incl. contractor inception results
 Stakeholder consultation plan, activities and timing
 Draft OPC questionnaire (launch of written consultation)
21/10/2021  Progress on evaluation/IA incl. contractor interim results
 Update on stakeholder consultation activities
27/01/2022  Contractor feedback on final workshop results
 Complete draft SWD on evaluation for endorsement
 Progress draft SWD on impact assessment
 Progress draft SWD on consultation synopsis
10/02/2022  Complete draft SWD on impact assessment for endorsement
 Complete draft SWD on consultation synopsis for endorsement
16/06/2022  Revised draft SWD on impact assessment for endorsement of
changes addressing the RSB opinion
105
https://intragate.ec.europa.eu/decide/sep/?view-dossier-details-id=DORSALE-DOSSIER-2021-5573
106
SG, SJ, AGRI, BUDG, EAC, ECFIN, EMPL, ENER, HOME, INTPA, JRC, JUST, NEAR, REGIO, RTD, SANTE.
65
12. 3. CONSULTATION OF THE RSB
The RSB was consulted on this impact assessment at a meeting on 16 March 2022.
13. 4. EVIDENCE, SOURCES AND QUALITY
Evidence and sources
Evidence Sources
Desk research  Statistical data and metadata currently published (baseline)
 Legal acts and explanatory memoranda related to the
current legislation
 Commission reports on implementation of the current
legislation
 Methodological guidelines and papers
 International recommendations
 Policy documents establishing statistical needs
A comprehensive list of documents reviewed is provided in
Table 12.
Opinion of statistics users:
Commission services
 Written consultation with the Commission network of
statistical correspondents
 Bilateral exchanges to pinpoint specific needs
 OPC survey
Opinion of other statistics
users
 Topical workshop with selected organisational statistics
users on problem definition
 In-depth interviews with selected organisational statistics
users
 OPC survey
Opinion of statistics
producers
 Regular consultation of relevant Commission expert
groups (see below)
 Case studies with five selected Member States
 OPC survey
 Targeted survey with NSIs complementing the OPC
Expert advice used
To seek advice and inputs on the progress of evaluation and impact assessment, Eurostat has
engaged the following Expert Groups (see Register of Commission Expert Groups107
) regularly:
 Working Group on Population and Housing Censuses (E01544) and its subgroup the Task Force on
the Future of Censuses;
 Working Group on Population Statistics (E03076);
 European Directors of Social Statistics (E01552).
107
https://ec.europa.eu/transparency/expert-groups-register/screen/home
66
The European Statistical System Committee was also informed about the progress.
External support study
Eurostat carried out this impact assessment with support from a contractor study contributed by ICF
SA, Belgium. In particular, the quantitative cost analysis, the methodological approach to scoring
and ranking of the policy options as well as support on stakeholder consultation activities, were
provided through the support study. Parts of this impact assessment SWD are therefore based on
the final report on the study supporting the impact assessment and other analysis documents
prepared by the contractor.
Quality
Based on the evidence sources and expert advice mentioned, Eurostat has carried out this impact
assessment using various inputs from an external support provider on stakeholder consultation, cost
quantification and options design and comparison as mentioned above. Eurostat has monitored the
work of the external support contractor regularly (at least every two weeks) and assessed the quality
of the final report on impact assessment. The overall work quality and deliverables were found to
be in line with the contract and generally sufficient to be used for this evaluation [preliminary draft
as contract still runs until 7 April 2022; revisit after final acceptance of all contract deliverables].
All external references relevant for aspects of the impact assessment were also added to this report.
67
Table 12 – List of reviewed documents
Author Published Title
Agilis 2017 Analysis of the legal and
institutional environment in the EU
Member States and EFTA Countries
DG ECFIN 2009 – 2021 Ageing Report 2009, 2012, 2015, 2018, 2021
DG ECFIN 2021 Euro Area Housing Markets: Trends, Challenges and
Policy Responses
DG HOME 2009 – 2020 Annual reports on migration and asylum
DG JUST / Subgroup on
Equality Data
2021 Guidance note on the collection and use of equality data
based on racial or ethnic origin
DG JUST 2018 Guidelines on improving the collection and use of
equality data
DG JUST 2017 Analysis and comparative review of equality data
collection practices in the European Union - Equality
data indicators: Methodological approach Overview per
EU Member State Technical annex
DG JUST 2017 Legal framework and practice in the EU Member States
DG JUST 2017 Data collection in relation to LGBTI People
DG JUST 2017 Data collection in the field of ethnicity
DG JUST 2016 European handbook on equality data
DG JUST 2015 Special Eurobarometer 437 – Discrimination in the EU in
2015
DG REGIO 2022 8th
Report on Economic, Social and Territorial Cohesion
DG REGIO 2017 7th
Report on Economic, Social and Territorial Cohesion
DG REGIO 2014 6th
Report on Economic, Social and Territorial Cohesion
ECB 2018 The state of the housing market in the euro area (ECB
Economic Bulletin, Issue 7/2018)
European Commission 2021 GREEN PAPER ON AGEING Fostering solidarity and
responsibility between generations
European Commission 2021 REPORT FROM THE COMMISSION TO THE
EUROPEAN PARLIAMENT AND THE COUNCIL on
the implementation of Regulation (EC) No 862/2007
European Commission 2018 COMMISSION IMPLEMENTING REGULATION
(EU) 2018/1799 on a temporary direct statistical action
for the dissemination of selected topics of the 2021
population and housing census geocoded to a 1 km2
grid
European Commission 2018 REPORT FROM THE COMMISSION TO THE
EUROPEAN PARLIAMENT AND THE COUNCIL on
the implementation of Regulation (EC) No 862/2007
European Commission 2018 REPORT FROM THE COMMISSION TO THE
EUROPEAN PARLIAMENT AND THE COUNCIL on
the implementation of Regulation (EU) No 1260/2013
European Commission 2017 COMMISSION IMPLEMENTING REGULATION
(EU) 2017/881 implementing Regulation (EC) No
68
Author Published Title
763/2008 on population and housing censuses, as regards
the modalities and structure of the quality reports and the
technical format for data transmission
European Commission 2017 COMMISSION REGULATION (EU) 2017/712
establishing the reference year and the programme of the
statistical data and metadata for population and housing
censuses provided for by Regulation (EC) No 763/2008
European Commission 2017 COMMISSION IMPLEMENTING REGULATION
(EU) 2017/543 laying down rules for the application of
Regulation (EC) No 763/2008 on population and housing
censuses as regards the technical specifications of the
topics and of their breakdowns
European Commission 2015 REPORT FROM THE COMMISSION TO THE
EUROPEAN PARLIAMENT AND THE COUNCIL on
the implementation of Regulation (EC) No 862/2007
European Commission 2014 COMMISSION IMPLEMENTING REGULATION
(EU) No 205/2014 laying down uniformed conditions for
the implementation of Regulation (EU) No 1260/2013 on
European demographic statistics, as regards breakdowns
of data, deadlines and data revisions
European Commission 2012 REPORT FROM THE COMMISSION TO THE
EUROPEAN PARLIAMENT AND THE COUNCIL on
the implementation of Regulation (EC) No 862/2007
European Commission 2011 Proposal for a REGULATION OF THE EUROPEAN
PARLIAMENT AND OF THE COUNCIL on European
statistics on demography
European Commission 2010 COMMISSION REGULATION (EU) No 1151/2010
implementing Regulation (EC) No 763/2008 on
population and housing censuses, as regards the
modalities and structure of the quality reports and the
technical format for data transmission
European Commission 2010 COMMISSION REGULATION (EU) No 519/2010
adopting the programme of the statistical data and of the
metadata for population and housing censuses provided
for by Regulation (EC) No 763/2008
European Commission 2010 COMMISSION REGULATION (EU) No 351/2010
implementing Regulation (EC) No 862/2007 on
Community statistics on migration and international
protection as regards the definitions of the categories of
the groups of country of birth, groups of country of
previous usual residence, groups of country of next usual
residence and groups of citizenship
European Commission 2009 COMMISSION REGULATION (EC) No 1201/2009
implementing Regulation (EC) No 763/2008 on
population and housing censuses as regards the technical
specifications of the topics and of their breakdowns
European Commission 2007 Proposal for a Regulation of the European Parliament and
of the Council on population and housing censuses
69
Author Published Title
European Commission 2005 Proposal for a Regulation of the European Parliament and
of the Council on Community statistics on migration and
international protection
European Committee of the
Regions
2016 The impact of demographic change on European regions
European Parliament 2021 Resolution of 21 January 2021 on access to decent and
affordable housing for all
European Parliament 2019 Demographic trends in EU regions
European Parliament and
Council
2019 REGULATION (EU) 2019/1700 establishing a common
framework for European statistics relating to persons and
households, based on data at individual level collected
from samples
European Parliament and
Council
2013 REGULATION (EU) No 1260/2013 on European
demographic statistics
European Parliament and
Council
2013 REGULATION (EU) No 549/2013 on the European
system of national and regional accounts in the European
Union
European Parliament and
Council
2009 REGULATION (EC) No 223/2009 on European statistics
European Parliament and
Council
2008 REGULATION (EC) No 763/2008 on population and
housing censuses
European Parliament and
Council
2007 REGULATION (EC) No 862/2007 on Community
statistics on migration and international protection
European Parliamentary
Research Service (EPRS)
2021 Demographic Outlook for the European Union
European Parliamentary
Research Service (EPRS)
2013 How can regional and cohesion policies tackle
demographic challenges?
European Commission’s
Joint Research Centre
2022 Data Innovation in Demography, Migration and Human
Mobility
Eurostat 2020 European statistical system handbook for quality and
metadata reports
Eurostat 2007 – 2021 Sustainable development in the European Union —
Monitoring reports on progress towards the SDGs in an
EU context 2007, 2011, 2015, 2019, 2021
Eurostat 2021 The European System of Accounts — ESA 2010 —
interactive version
Eurostat 2020 Quality assurance framework of the European statistical
system
Eurostat 2010 – 2020 Report on the impact of demographic change 2010, 2015,
2020
Eurostat 2014 – 2020 Eurostat Regional Yearbook 2010, 2014, 2020
Eurostat 2009 – 2020 Eurostat User Satisfaction Survey reports 2009, 2013 –
2017, 2019 – 2020
Eurostat 2019 EU legislation on the 2021 population and housing
70
Author Published Title
censuses – explanatory notes
Eurostat 2018 European statistics Code of Practice for the national
statistical authorities and Eurostat
Eurostat 2011 EU legislation on the 2011 population and housing
censuses – explanatory Notes
ESSnet KOMUSO 2019 Quality Guidelines for Multisource Statistics
ESSnet KOMUSO 2019 Quality Guidelines on Frames for Social Statistics
ICF 2022 Final Report on impact assessment support study for
European statistics on population
ICF 2021 Inception Report on support study for European statistics
on population
United Nations 2017 Principles and Recommendations for Population and
Housing Censuses
UNECE 2018 Guidelines on the use of registers and administrative data
for population and housing censuses
UNECE 2015 Recommendations for the 2020 Censuses of Population
and Housing
UNECE 2014 Measuring population and housing. Practices of UNECE
countries in the 2010 round of censuses
UNECE 2008 Measuring population and housing Practices of UNECE
countries in the 2000 round of censuses
71
ANNEX 2: WHO IS AFFECTED AND HOW?
14. 1. PRACTICAL IMPLICATIONS OF THE INITIATIVE
14.1. 1.1. Detailed quantitative costs by policy option
This part of Annex 2 provides detailed quantitative cost assessments of all options including the
baseline (option A), based on a detailed cost itemisation applied in the external support study
(footnote 71; see also Annex 3) and mapped onto BRG cost items in Table 13 below. Incremental
costs of options B – D are aligned to the main option characteristics introduced in section 4.2, and
all amounts are given in thousands of 2021 EUR with a range of +/-5% around the best estimate
derived from the cost model to account for a margin of error.108
As explained further in Annex 3,
one-off costs are calculated over 3 years to account for the transition period to a new legal
framework.
108
The final report on the support study (footnote 71) provides more extensive variations including additional error
margins of +/-10% and +/-20% around the best estimate.
72
Table 13 – Mapping between BRG cost items and detailed cost items assessed in this study, including definitions
BRG cost
items
One-off /
recurrent
Stake-
holders
Detailed cost
items
Definitions
Direct
adjustment
costs
One-off Eurostat Regulatory costs Costs required to prepare and draft the new
regulation, including adopting executive acts,
communication and provision of training
One-off MS/NSIs Administrative
costs - training,
guidance and
support
Resources required to develop, implement and
communicate any training required and/or
commission experts to cover skills gaps
Direct
administrative
costs
Recurrent MS/NSIs Administrative
costs - design and
implementation of
new procedures
Resources required to design and implement a
programme to collect, accumulate, manage,
clean and transmit demographic data (in the
manner required by the regulation) including
through collaboration/ data transmission
between NSIs and national authorities
Direct
enforcement
costs
Recurrent Eurostat Enforcement costs
- monitoring and
reporting
Resources required to monitor MS alignment
to, and compliance with the requirements of
the new regulation (including around data
protection, data quality and transparency) and
to receive and publish data provided
Recurrent NSIs/MS Enforcement costs
- monitoring and
reporting
Resources required to monitor and ensure
compliance to the regulation in the MS
Recurrent NSIs/MS Compliance costs Resources required to ensure compliance with
requirements, including adopting national
arrangements or legislation to ensure
cooperation with relevant source data holders
IT costs One-off /
recurrent
Eurostat
NSIs/MS
IT costs New software/application/IT update necessary
for implementing the new measure
Direct
regulatory fees
and charges
Not
estimated
Indirect costs Not
estimated
73
14.1.1. 1.1.1.
14.1.2. 1.1.1. A – Baseline
The baseline scenario is described in section 6.1.1. The following tables present an overview of the
costs entailed by this policy option.
Table 14 – Overview of baseline costs (Detailed format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over 10
years)
One-off Recurrent (over 10
years)
A –
BASELINE
Limited
harmonisation,
separate
statistical
processes and
current
statistical
outputs
Population
frames or
registers
N/A N/A N/A 22,706 – 23,901
Population
Definitions
N/A N/A N/A 4,299 – 4,525
Mandatory
demography
statistics
N/A 2,377 – 2,502 N/A 89,739 – 94,462
Mandatory
migration
statistics
N/A 2,181 – 2,296 N/A 92,081 – 96,927
Mandatory
census
statistics
N/A 8,214 – 8,646 N/A 1,461,036 – 1,537,933
Voluntary
statistics
N/A 758 – 798 N/A 3,357 – 3,534
Cost savings N/A N/A N/A 111,730 – 123,492
Total baseline
costs
N/A 12,886 – 14,242 1,481,810 – 1,637,790
Table 15 – Overview of baseline costs (BRG format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over 10
years)
One-off Recurrent (over 10
years)
A –
BASELINE
Limited
harmonisation,
separate
statistical
processes and
current
statistical
outputs
Direct
adjustment costs
N/A N/A N/A N/A
Direct
administrative
costs
N/A N/A N/A 1,481,810 – 1,637,790
Direct
regulatory fees
and charges
N/A N/A N/A N/A
Direct
enforcement
N/A 12,886 – 14,242 N/A N/A
74
costs
Indirect costs N/A N/A N/A N/A
14.1.3. 1.1.2. Policy option B.1 – Limited harmonisation, separate statistical
processes, limited upgrade of statistical outputs
In policy option B.1, the status quo remains for the harmonisation of population definition
(limited), as well as for the statistical processes (separate resp. up to Member States).
The incremental costs of this option, relatively to the baseline, are incurred by the limited upgrade
of the statistical outputs. Updated statistical requirements are introduced in demography, migration,
and housing census statistics.
Under this Option, the European Commission would incur the following incremental costs:
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: EUR 97,116
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 97,072
o IT costs: N/A
Under this Option, all 27 Member states (NSIs) would incur the following incremental costs:
 One-off costs:
o Administrative costs: EUR 2,111,179
o IT costs: EUR 3,704,262
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 1,519,285
o Enforcement costs: EUR 360,945
o Administrative costs: EUR 1,450,227
o IT costs: EUR 3,635,085
75
Table 16 – Overview of incremental costs of policy option B.1 (Detailed format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over
10 years)
One-off Recurrent (over 10
years)
POLICY
OPTION B.1
Limited
harmonisation,
separate
statistical
processes,
limited update
of statistical
outputs
Regulatory costs 93 – 103 N/A N/A N/A
Administrative
costs - training,
guidance and
support
N/A N/A 2,006 – 2,217 N/A
Compliance costs N/A N/A N/A 1,443 – 1,595
Enforcement
costs - monitoring
and reporting
N/A 92 – 102 N/A 343 – 379
Administrative
costs - design and
implementation
of new
procedures
N/A N/A N/A 1,378 – 1,523
IT costs 92 – 102 0 3,519 – 3,889 3,453 – 3,817
Table 17 – Overview of incremental costs of policy option B.1 (BRG format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent
(over 10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION B.1
Limited
harmonisation,
separate
statistical
processes,
limited update
of statistical
outputs
Direct adjustment
costs
93 – 103 N/A 2,006 – 2,217 N/A
Direct
administrative costs
N/A N/A N/A 1,378 – 1,523
Direct enforcement
costs
N/A 92 – 102 N/A 1,786 – 1,974
IT costs 92 – 102 0 3,519 – 3,889 3,453 – 3,817
Direct regulatory
fees and charges
Not estimated Not estimated Not estimated Not estimated
Indirect costs Not estimated Not estimated Not estimated Not estimated
76
14.1.4. 1.1.3. Policy option B.2 - Limited harmonisation, separate statistical
processes, more expansive upgrade of statistical outputs and flexibility
In policy option B.2, similarly to Option B.1, the status quo remains for the harmonisation of
population definition (limited), as well as for the statistical processes (separate resp. up to Member
States). However, the more expansive upgrade of statistical outputs can be expected to accelerate
indirectly the current transition towards integrated register-based systems at national levels.
The incremental costs of this option, relatively to the baseline, are incurred by an update of the
statistical outputs (more expansive update of requirements for census, demography, and migration
statistics compared to option B.1 and Option C.1) which also includes to make mandatory some of
the existing voluntary statistics data.
Under policy option B.2, the European Commission would incur the following incremental costs
due to the more important upgrade of statistical outputs:
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: EUR 288,602
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 304,618
o IT costs: EUR 65,425
All 27 Member states (NSIs) would incur the following incremental costs due to the more
important upgrade of statistical outputs:
 One-off costs:
o Administrative costs: EUR 6,929,358
o IT costs: EUR 10,841,743
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 9,668,179
o Enforcement costs: EUR 2,381,134
o Administrative costs: EUR 9,087,189
o IT costs: EUR 12,839,075
77
Table 18 – Overview of incremental costs of policy option B.2 (Detailed format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over
10 years)
One-off Recurrent (over 10
years)
POLICY
OPTION B.2
Limited
harmonisation,
separate
statistical
processes,
upgrade of
statistical
outputs and
flexibility
Regulatory costs 93 - 103 N/A N/A N/A
Administrative
costs - training,
guidance and
support
N/A N/A 6,583 - 7,276
Compliance
costs
N/A
Enforcement
costs -
monitoring and
reporting
N/A N/A N/A 9,185 – 10,152
Administrative
costs - design
and
implementation
of new
procedures
N/A 289 – 320 N/A 2,262 – 2,500
IT costs N/A N/A N/A 8,633 – 9,542
Table 19 – Overview of incremental costs of policy option B.2 (BRG format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent
(over 10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION B.2
Limited
harmonisation,
separate
statistical
processes,
upgrade of
statistical
outputs and
flexibility
Direct adjustment
costs
93 – 103 N/A 6,583 – 7,276 N/A
Direct
administrative
costs
N/A N/A N/A 8,633 – 9,542
Direct
enforcement costs
N/A 289 – 320 N/A 11,447 – 12,652
IT costs 274 – 303 62 – 69 10,300 –11,384 12,197 – 13,481
Direct regulatory
fees and charges
Not estimated Not estimated Not estimated Not estimated
Indirect costs Not estimated Not estimated Not estimated Not estimated
78
14.1.5. 1.1.4. Policy option C.1 - Improved harmonisation, separate statistical
processes, limited upgrade of statistical outputs
In policy option C.1, the status quo remains with regards to the statistical processes.
An improvement of the harmonisation of population definitions is however introduced.
The incremental costs associated with a better harmonisation of population definition are estimated
to be the following for the European Commission (Eurostat):
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: N/A
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 513,322
o IT costs: N/A
The incremental costs incurred by Member States (NSIs)109
to meet the requirements of an
improved harmonisation of population definitions are estimated to be the following:
 One-off costs:
o Administrative costs: EUR 2,627,940
o IT costs: EUR 4,015,460
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 6,241,425
o Enforcement costs: EUR 6,241,425
o Administrative costs: EUR 6,510,559
o IT costs: N/A
In terms of statistical outputs, the incremental costs of this option, relatively to the baseline, are
incurred by the limited upgrade of these. Similarly to Option B.1, updated statistical requirements
are introduced in demography, migration and housing census statistics.
Under this Option the European Commission would incur the following incremental costs due to
the limited upgrade of statistical outputs:
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: EUR 97,116
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 97,072
o IT costs: N/A
Under this Option, all 27 Member states (NSIs) would incur the following incremental costs:
109
It was assumed that the Member States that replied to the NSIs survey question ‘Which population definition(s) is
used in your NSI?’ with ‘Combined’ or other’ would be most likely to transition to the use of a population definition
based on the strict usual residence concept, should the regulation restrict the extent to which NSIs can be exempt from
using this population definition. Those NSIs are: AT, CZ, DE, EE, ES, LV, PL, SI.
79
 One-off costs:
o Administrative costs: EUR 2,111,179
o IT costs: EUR 3,704,262
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 1,519,285
o Enforcement costs: EUR 360,945
o Administrative costs: EUR 1,450,227
o IT costs: EUR 3,635,085
80
Table 20 – Overview of incremental costs of policy option C.1 (Detailed format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over 10
years)
One-off Recurrent (over
10 years)
POLICY
OPTION C.1
Improved
harmonisation,
separate
statistical
processes,
limited update
of statistical
outputs
Regulatory costs 186 – 205 N/A N/A N/A
Administrative
costs - training,
guidance and
support
N/A N/A 4,502 – 4,976 N/A
Compliance
costs
N/A N/A N/A 7,373 – 8,149
Enforcement
costs -
monitoring and
reporting
N/A 582 – 643 N/A 6,272 – 6,932
Administrative
costs - design
and
implementation
of new
procedures
N/A N/A N/A 7,563 – 8,359
IT costs 92 – 102 0 7,334 – 8,106 3,453 – 2,817
Table 21 – Overview of incremental costs of policy option C.1 (BRG format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over
10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION C.1
Improved
harmonisation,
separate
statistical
processes,
limited update
of statistical
outputs
Direct adjustment
costs
186 – 205 N/A 4,502 – 4,976 N/A
Direct
administrative
costs
N/A N/A N/A 7,563 – 8,359
Direct
enforcement costs
N/A 582 – 643 N/A 13,645 – 15,081
IT costs 92 – 102 0 7,334 – 8,106 3,454 – 3,817
Direct regulatory
fees and charges
Not estimated Not estimated Not estimated Not estimated
Indirect costs Not estimated Not estimated Not estimated Not estimated
81
14.1.6. 1.1.5. Policy option C.2 – Improved harmonisation, separate statistical
processes, more expansive upgrade of statistical outputs and flexibility
In policy option C.2, the status quo remains with regards to the statistical processes. However, the
more expansive upgrade of statistical outputs can be expected to accelerate indirectly the current
transition towards integrated register-based systems at national levels.
An improvement of the harmonisation of population definitions is introduced, similarly to
Option C.1. Therefore, the costs incurred by both the European Commission (Eurostat) and NSIs
for harmonising better population definitions are the same than those estimated in policy option
C.1.
The incremental costs associated with a better harmonisation of population definition are estimated
to be the following for the European Commission (Eurostat):
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: N/A
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 513,322
o IT costs: N/A
The incremental costs incurred by NSIs to meet the requirements of improved harmonisation of
population definitions are estimated to be:
 One-off costs:
o Administrative costs: EUR 2,627,940
o IT costs: EUR 4,015,460
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 6,241,425
o Enforcement costs: EUR 6,241,425
o Administrative costs: EUR 6,510,559
o IT costs: N/A
The incremental costs of this option C.2, relatively to the baseline, are also incurred by an update
of the statistical outputs (more expansive update of requirements for census, demography, and
migration statistics compared to option B.1 and Option C.1) which also includes to make
mandatory some of the existing voluntary statistics data. These costs incurred by the more
expensive upgrade of statistical outputs are the same than those estimated in Option B.2.
Under policy option C.2, the European Commission would incur the following incremental costs
due to the more important upgrade of statistical outputs:
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: EUR 288,602
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 304,618
o IT costs: EUR 65.425
All EU-27 (NSIs) would incur the following incremental costs due to the more important upgrade
of statistical outputs:
82
 One-off costs:
o Administrative costs: EUR 6,929,358
o IT costs: EUR 10,841,743
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 9,668,179
o Enforcement costs: EUR 2,381,134
o Administrative costs: EUR 9,087,189
o IT costs: EUR 12,839,075
83
Table 22 – Overview of incremental costs of policy option C.2 (Detailed format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over
10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION C.2
Improved
harmonisation,
separate
statistical
processes,
upgrade of
statistical
outputs and
flexibility
Regulatory costs 186 – 205 N/A N/A N/A
Administrative
costs – training,
guidance and
support
N/A N/A 9,079 – 10,035 N/A
Compliance
costs
N/A N/A N/A 15,114 – 16,705
Enforcement
costs –
monitoring and
reporting
N/A 779 – 861 N/A 8,191 – 9,054
Administrative
costs – design
and
implementation
of new
procedures
N/A N/A N/A 14,818 – 16,378
IT costs
274 – 303 62 – 69
14,114 –
15,600
12,197 – 13,481
Table 23 – Overview of incremental costs of policy option C.2 – BRG format
European Commission NSIs/Member States – all 27
One-off Recurrent
(over 10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION C.2
Improved
harmonisation,
separate
statistical
processes,
upgrade of
statistical
outputs and
flexibility
Direct adjustment
costs
186 - 205 N/A 9,079 – 10,035 N/A
Direct
administrative
costs
N/A N/A N/A
14,818 – 16,378
Direct
enforcement costs
N/A N/A N/A 23,305 – 25,759
IT costs 274 - 303
62- 69 14,114 –
15,600
12,197 – 13,481
Direct regulatory
fees and charges
Not estimated Not estimated Not estimated Not estimated
Indirect costs Not estimated Not estimated Not estimated Not estimated
84
14.1.7. 1.1.6. Policy option D.1 – Full harmonisation, separate statistical
processes, major upgrade of statistical outputs
In policy option D.1, statistical processes remain separated. This option however introduces the full
harmonisation of population definitions based on the usual residence concept, as well as a major
upgrade of statistical outputs. Similar to Option B.2, requirements are more important compared to
those in policy option B.2 and C.2, covering existing and new topics of demography, migration, and
population and housing census statistics. Some of census outputs will be available more frequently
than every 10 years. With regards to the full harmonisation of population definitions, the
incremental costs to the Commission (Eurostat) associated with this measure are estimated as:
 One-off costs:
o Regulatory costs: EUR 97,658
o IT costs: N/A
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 515,322
o IT costs: N/A
The full harmonisation of population definitions will incur incremental costs to Member States.
NSIs which are currently not using the definition based on the strict usual residence concept to
define their population are expected to incur the following incremental costs:
 One-off costs:
o Administrative costs: EUR 7,470,299
o IT costs: EUR 8,532,8541
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 4,949,187
o Enforcement costs: EUR 4,949,187
o Administrative costs: EUR 14,661,584
o IT costs: N/A
Regarding the major upgrade of statistical outputs, the following incremental costs are expected
to be incurred by the Commission (Eurostat):
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: EUR 388,463
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 514.821
o IT costs: EUR 151.197
It was estimated that all 27 Member states (NSIs) would incur the following incremental costs due
to the more detailed and new requirements in terms of statistical outputs:
 One-off costs:
o Administrative costs: EUR 11,400,365
o IT costs: EUR 14,003,918
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 14,916,619
o Enforcement costs: EUR 4,139,822
o Administrative costs: EUR 16,043,674
o IT costs: EUR 15,942,196
85
Table 24 – Overview of incremental costs of policy option D.1 (Detailed format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over
10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION D.1
Full
harmonisation,
separate
statistical
processes,
major upgrade
of statistical
outputs
Regulatory costs 186 – 205 N/A N/A N/A
Administrative
costs - training,
guidance and
support
N/A N/A
17,972 –
19,814
N/A
Compliance
costs
N/A N/A N/A 18,873 – 20,859
Enforcement
costs -
monitoring and
reporting
N/A 979 – 1,082 N/A 8,635 – 9,543
Administrative
costs - design
and
implementation
of new
procedures
N/A N/A N/A 29,170 – 32,241
IT costs
369 – 408 144 – 159
21,410 –
23,664
15,145 – 16,739
Table 25 – Overview of incremental costs of policy option D.1 (BRG format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent
(over 10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION D.1
Full
harmonisation,
separate
statistical
processes,
major upgrade
of statistical
outputs
Direct adjustment
costs
186 – 205 N/A
17,972 –
19,814
N/A
Direct
administrative
costs
N/A N/A N/A 29,170 – 32,241
Direct
enforcement costs
N/A 979 – 1,082 N/A 27,508 – 30,402
IT costs 369 – 408 144 – 159
21,410 –
23,664
15,145 – 16,739
Direct regulatory
fees and charges
Not estimated Not estimated Not estimated Not estimated
Indirect costs Not estimated Not estimated Not estimated Not estimated
86
14.1.8. 1.1.7. Policy option D.2 – Full harmonisation, redeveloped and integrated
statistical processes in a single and flexible framework, major upgrade of
statistical outputs
Policy option D.2 introduces the development of integrated statistical processes in a single and
flexible framework. Under this option, NSIs are required to set up national statistical registers. It
also introduces the full harmonisation of population definitions based on the usual residence
concept. There is also a major upgrade of statistical outputs (requirements are more important
than those in policy options B.2 and C.2), covering existing and new topics of demography,
migration, and population and housing census statistics. Some of census outputs will be available
more frequently than every 10 years.
For the implementation of national statistical registers, the incremental costs for the European
Commission (Eurostat) are estimated to be:
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: EUR 145,673
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 172,573
o IT costs: EUR 476,510
This measure, specific to policy option D.2, would also incur incremental costs for all Member
States (NSIs) without a national statistical register in place or planned110
. The incremental costs
related to the implementation of national statistical registers are estimated to be:
 One-off costs:
o Administrative costs: EUR 7,401,716
o IT costs: EUR 3,011,595
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 27,623,368
o Enforcement costs: EUR 19,796,747
o Administrative costs: EUR 4,842,327
o IT costs: EUR 1,970,236
With regard to the full harmonisation of population definitions, the incremental costs to the
Commission (Eurostat) under policy option D.2 are estimated to be:
 One-off costs:
o Regulatory costs: EUR 97,658
o IT costs: N/A
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 512,322
o IT costs: N/A
110
NSI survey responses (Q. ‘Does your NSI maintain a national statistical register?’ The following responded ‘No
and not planned’: Germany, Ireland, Luxembourg, Poland, Romania, Slovenia).
87
Full harmonisation of population definitions will incur incremental costs to Member States. NSIs
not currently using the definition based on the strict usual residence concept to define their
population are expected to incur the following incremental costs:
 One-off costs:
o Administrative costs: EUR 7,470,299
o IT costs: EUR 8,532,854
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 4,949,187
o Enforcement costs: EUR 4,989,187
o Administrative costs: EUR 14,661,584
o IT costs: N/A
For the major upgrade of statistical outputs, the following incremental costs are expected to be
incurred by the Commission (Eurostat):
 One-off costs:
o Regulatory costs: EUR 97,698
o IT costs: EUR 388,463
 Recurrent (over 10 years) costs:
o Enforcement costs: EUR 514,821
o IT costs: EUR 151,197
EU-27 (NSIs) would incur the following incremental costs, due to the more detailed and new
requirements in terms of statistical outputs:
 One-off costs:
o Administrative costs: EUR 11,400,365
o IT costs: EUR 14,003,918
 Recurrent (over 10 years) costs:
o Compliance costs: EUR 14,916,619
o Enforcement costs: EUR 4,139,822
o Administrative costs: EUR 16,043,674
o IT costs: EUR 15,942196
88
Table 26 – Overview of incremental costs of policy option D.2 (Detailed format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent (over
10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION D.2
Full
harmonisation,
redeveloped
and integrated
statistical
processes in a
single and
flexible
framework,
major upgrade
of statistical
outputs
Regulatory costs 278 – 308 N/A N/A N/A
Administrative
costs - training,
guidance and
support
N/A N/A
24,959 –
27,586
N/A
Compliance
costs
N/A N/A N/A 45,115 – 49,864
Enforcement
costs -
monitoring and
reporting
N/A 1,143 – 1,263 N/A 27,441 – 30,330
Administrative
costs - design
and
implementation
of new
procedures
N/A N/A N/A 33,770 – 37,325
IT costs
507 – 561 596 – 659
24,271 –
26,826
17,017 – 18,808
Table 27 – Overview of incremental costs of policy option D.2 (BRG format in thousands of 2021 EUR)
European Commission NSIs/Member States – all 27
One-off Recurrent
(over 10 years)
One-off Recurrent (over
10 years)
POLICY
OPTION D.2
Full
harmonisation,
redeveloped
and integrated
statistical
processes in a
single and
flexible
framework,
major upgrade
of statistical
outputs
Direct adjustment
costs
278 – 308 N/A 24,959 - 27,586 N/A
Direct
administrative
costs
N/A N/A N/A 33,770 – 37,325
Direct
enforcement costs
N/A 1,143 – 1,263 N/A 72,556 – 80,194
IT costs
507 – 561 596 - 659 24,271 –
26,826
17,017 – 18,808
Direct regulatory
fees and charges
Not estimated Not estimated Not estimated Not estimated
Indirect costs Not estimated Not estimated Not estimated Not estimated
89
14.2. 1.2. Detailed qualitative scoring by policy option
14.2.1. 1.2.1. A – Baseline
Assessment criterion Score Brief description
Consistency with
proportionality principle
(Pass /
Uncertain /
Fail)
Does the option go
beyond what is necessary
to satisfactorily achieve
the objectives?
Pass Under the baseline, no further EU action would be taken. There is no
risk that the option will go beyond what is necessary to achieve the
policy objectives.
Is the scope of the option
limited to those aspects
that Member States
cannot achieve
satisfactorily on their
own, and where the
Union can do better?
Pass Under the baseline, no further EU action would be taken. Current EU
action is limited to those aspects that Member States cannot achieve
satisfactorily on their own, such as statistical population frames
fulfilling certain common criteria. Member States have the flexibility to
use a national population definition and to provide additional data on a
voluntary basis. Results from the evaluation show that current EU
action provides added value.
Are costs for the Union,
national governments,
regional or local
authorities, economic
operators or citizens
commensurate with the
objectives of the
initiative?
Pass Under the baseline, no further EU action would be taken. Thus, costs
for data producers would be minimised.
However, data users will incur costs because the population base is not
harmonised, for example costs associated with obtaining data directly
from NSIs and additional data processing.
Is the form of action
(choice of instrument) as
simple as possible and
coherent with
satisfactory achievement
of the objective and
effective enforcement?
Pass Under the baseline, no further EU action would be taken. Each
statistical domain would continue to be implemented as a separate
statistical process based on its own EU legislation. Thus, a patchwork
of legislation would remain. The choice of action may not be as simple
as possible.
Effectiveness in
achieving policy
objectives:
(-4 to +4)
SO1:
To ensure complete,
coherent and comparable
European population
statistics
0 Under the baseline, no further EU action would be taken to improve
completeness, comparability or coherence of ESOP.
The baseline fails to meet SO1.
SO2:
To ensure the availability
of timely and frequent
0 Under the baseline, no further EU action would be taken to improve
timeliness and frequency of statistics outputs.
90
population statistics to
meet users’ needs
The baseline fails to meet SO2.
SO3:
To provide statistics that
are sufficiently
comprehensive in terms
of relevant topics and
sufficiently detailed in
terms of characteristics
and breakdowns
0 Under the baseline, no further EU action would be taken to enhance the
detail and scope of population statistics in terms of characteristics and
breakdowns.
The baseline fails to meet SO3.
SO4:
To promote frameworks
that are sufficiently
flexible to adapt to
evolving policy needs
and opportunities from
new sources
0 Under the baseline, no further EU action would be taken and no new
frameworks to ensure flexibility would be proposed.
The baseline fails to meet SO4.
Impacts on fundamental
rights
0
All criteria were scored ‘0’ at the baseline to set a meaningful reference
for the subsequent policy options. Nonetheless, stakeholder
consultations revealed shortcomings in European statistics in relation to
fundamental rights protection. Several respondents highlighted the lack
of disaggregated data across population subgroups at risk of inequality
or discrimination, including disaggregated and comparable data on all
grounds in Article 21 of the Charter111
.
The lack of a legal obligation of equality data collection on all Charter
grounds was noted as a factor preventing systematic collection of data
to regularly assess the situation of individuals and groups of individuals
at risk of inequality and discrimination. Equality bodies and public
institutions would benefit from such statistics in their decision-making.
The lack of equality data was highlighted in the EU Gender Equality
Strategy 2020-2025, the LGBTIQ Equality Strategy, the EU Roma
Strategic Framework and the EU Anti-racism Action Plan 2020-2025.
Social impacts 0
All criteria were scored ‘0’ at the baseline to set a meaningful reference
for the subsequent policy options.
Population statistics are the backbone of all social statistics as they
provide reference information on the entire population and its basic
demographic characteristics. Many of the proposed measures in
subsequent policy options are expected to contribute to modernising
social statistics, thus maintaining the status quo is suboptimal.
Modernisation of social statistics is needed to support various Union
actions, such as the promotion of social inclusion and the monitoring of
social cohesion at national and regional level. Demographic changes
111
Sex, race, colour, ethnic or social origin, genetic features, language, religion or belief, political or any other opinion,
membership of a national minority, property, birth, disability, age or sexual orientation.
91
such as ageing, increasing life expectancy, declining fertility and
depopulation of some areas have social and economic impacts.
European statistics on population were crucial for the Commission’s
report on the impact of demographic change and the Green Paper on
Ageing; population data are equally crucial for the Long-Term Vision
for Rural Areas initiative, to monitor the changes in the age structure of
the agricultural population, population density and rural access to
services.
Environmental impacts 0 All criteria were scored ‘0’ at the baseline to set a meaningful reference
for the subsequent policy options.
The European Green Deal envisages a new growth strategy that will
transform the Union into a modern, resource-efficient and competitive
economy where no person and no place is left behind. Improving the
population and housing data with sufficient regional detail, and in
particular grid data, is essential to understanding relevant issues such as
the spatial dimension of exposure to water, air and soil pollution or
access to smart mobility.
Coherence with
overarching objectives of
EU policy
(-4 to +4)
Coherence 0 Maintaining the status quo would mean that some shortcomings of the
current framework would not be resolved (e.g. lack of harmonisation
and flexibility, insufficient data to support the development of EU
legislation on politically relevant groups or topics, such as equality, the
Green Deal, rural population characteristics).
A ‘no action’ approach would have a somewhat negative effective on
effective legal coherence between the ESOP and new EU legal
instruments.
Efficiency
The baseline has a neutral overall effect on efficiency, as no
incremental impacts are incurred by definition of the baseline. All
stakeholders will continue to incur the current level of baseline costs.
Baseline benefits, including improved quality, timeliness and
availability of statistical outputs across the EU relative to the situation
in which data were provided voluntarily, will also be sustained.
Overall baseline efficiency may be relatively high – analysis from the
evaluation suggests that the current legal framework delivered benefits
efficiently.
Overall assessment
Under the baseline, there will be no specific policy changes at EU level and all measures represent a
continuation of the status quo. However, it is assumed the Demography Regulation will be amended to extend
its application.
This option may fail the proportionality principle if the legislative framework remains overly complex and
gives rise to an unnecessary cost burden for data users.
92
This option would not be effective in addressing the problems identified with the current EU legal framework.
Population statistics would continue to lack comparability and countries would continue to provide statistics
covering different characteristics and at different levels of spatial and temporal disaggregation. Data users,
including policymakers, would continue to lack access to the datasets and breakdowns they need, which may
(indirectly) negatively impact fundamental rights policymakers would lack the information required to support
Union actions to protect groups at risk of inequality and discrimination. It would also have indirect negative
social impacts, as modernisation of social statistics is needed to support Union actions, such as the promotion
of social inclusion and the monitoring of social cohesion at national and regional level. Demographic changes
such as ageing, increasing life expectancy, declining fertility and depopulation of some areas have social and
economic impacts.
In terms of coherence with other EU instruments, the results from the evaluation show that the policy
landscape has evolved considerably since the current legislation came into force. Several new synergies have
emerged in recent years due to an increased focus on policy areas such as non-discrimination and equality, the
Green Transition and digitalisation. As it stands, the current legal framework fails to capture characteristics
and details of politically relevant groups and topics.
Feedback from the public consultation revealed that the current legislative framework is perceived as enabling
the efficient production of statistics in a manner that generates greater value than burden. Given that the same
level of ongoing costs will continue to be incurred, the overall expected efficiency of this option is neutral,
assuming no net change.
In summary, maintaining the baseline would not be effective in addressing the problems identified with ESOP
framework, as population statistics would continue to lack comparability and the necessary disaggregation
required by policymakers and requested by other data users.
93
14.2.2. 1.2.2. Policy option B.1 - Limited harmonisation, separate statistical
processes, limited upgrade of statistical outputs
Assessment criterion Score Brief description
Consistency with
proportionality principle
(Pass /
Uncertain
/ Fall)
Does the option go
beyond what is
necessary to
satisfactorily achieve the
objectives?
Pass Policy option B.1 entails marginal changes compared to the baseline, and
the measures under this option do not go beyond what is necessary to
satisfactorily achieve the objectives.
Is the scope of the option
limited to those aspects
that Member States
cannot achieve
satisfactorily on their
own, and where the
Union can do better?
Pass Policy option B.1 would only marginally extend statistical requirements
compared to the baseline, and no new statistical topics would be
implemented. The scope of the option is limited to those aspects that
Member States cannot achieve satisfactorily on their own, and where the
Union can do better (e.g. updated statistical requirements and increased
quality of outputs).
Are costs for the Union,
national governments,
regional or local
authorities, economic
operators or citizens
commensurate with the
objectives of the
initiative?
Pass The expected costs for the Union, national governments, regional or local
authorities, economic operators or citizens, are expected to be minimal
when compared to the objectives to be achieved.
Is the form of action
(choice or instrument) as
simple as possible and
coherent with
satisfactory achievement
of the objective and
effective enforcement?
Pass The updated statistical requirements will be introduced in demography,
migration, and population and housing census statistics to meet user needs
on statistical topics covered by the current EU legislation, therefore the
form of action (choice of instrument) is as simple as possible, and
coherent with satisfactory achievement of the objective and effective
enforcement.
Effectiveness in
achieving the policy
objectives:
(-4 to +4)
SO1:
To ensure complete,
coherent and comparable
European population
statistics
1 In policy option B.1, similar to policy option B.2 and the baseline,
harmonisation of population definitions remains limited. NSIs would not
be required to implement national statistical registers for processing
population statistics. However, policy options B.1 and B.2 still improve
overall completeness and coherence relatively to the baseline (A).
By failing to harmonise the population definition (measure 2), there is a
risk that this option does not go far enough in improving the
comparability of European population statistics.
94
SO2:
To ensure the
availability of timely and
frequent population
statistics to meet users’
needs
1 In policy option B.1, legal deadlines for census outputs are shortened
(measure 9.2) and in topic 8 on timeliness, measure 8.2 ‘quality first’ is
chosen over 8.3 ‘timeliness first’.
Policy option B.1 does not improve the frequency of data availability
(measure 6.2 makes geo-referenced data mandatory every 10 years but
this is expected to impact more on comprehensiveness rather than
frequency of data availability). Policy option B.1 performs better in
meeting SO2 than the baseline (A), and equal to policy option C.1.
SO3:
To provide statistics that
are sufficiently
comprehensive in terms
of relevant topics and
sufficiently detailed in
terms of characteristics
and breakdowns
1 In policy option B.1 (and similar to policy option C.1), data on current
statistical output are more detailed than the baseline (more characteristics
on migration statistics; improved disaggregation of annual population
data and migration statistics). In terms of comprehensiveness of relevant
topics, policy option B.1 does not bring new statistical topics and
voluntary datasets and breakdowns are not made mandatory. Overall,
policy option B.1 (like policy option C.1) scores better than Policy the
baseline (A) but does not perform as well on SO3 as policy options B.2,
C.2 and D.1 and D.2.
Detail of statistics is increased slightly, but no new topics are added. By
omitting any changes under measures 12, 13 and 14, there is also a high
risk that this option will not address user needs sufficiently (e.g.
policymakers and NGOs concerned with equality).
SO4:
To promote frameworks
that are sufficiently
flexible to adapt to
evolving policy needs
and opportunities from
new sources
1 Policy option B.1 improves the framework flexibility compared to the
baseline by introducing limited flexibility to adapt outputs (measure 15.2)
as well as enabling better use of administrative and new sources for
statistical purposes (measures 17.2 and 17.3).
Impacts on fundamental
rights
0.5 Policy option B.1 represents a marginal improvement in fundamental
rights impacts. The ten measures under this option (3.2, 4.2, 6.2, 8.2, 9.2,
10.2, 11.2, 15.2, 17.2, 17.3) would collectively improve fundamental
rights only marginally and indirectly, as the measures are mostly aimed at
timeliness and detail of demographic statistics already collected. The
individual impact of each measure on fundamental rights is difficult to
establish.
Nevertheless, the collective improvement and modernisation of
population statistics envisaged is likely to have indirect impacts on
fundamental rights by improving the quality and availability of the
relevant data of specific vulnerable and at-risk groups. For example,
measure 10.2 ‘define migrant groups in more detail’ would help to better
understand the integration of third-country nationals and thereby inform
integration policies.
Social impacts 1 Policy option B.1 represents an improvement in social impact and, more
specifically, provision of social statistics for informed evidence-based
policy-making at EU and national level. The total of the seven measures
95
would collectively improve social impacts only marginally and indirectly,
as they are aimed at timeliness, harmonisation and quality of population
statistics collected.
Nevertheless, the collective improvement and modernisation of
population statistics envisaged is likely to have a slight indirect positive
impact by improving the quality and availability of the relevant data of
specific issues of interest to policy-making.
Environmental impacts 1 In the consultation, public authorities and NGOs highlighted that they rely
on population statistics for formulating and monitoring environmental
policies.
Policy option B.1 would have only a small positive environmental impact.
This impact would be indirect, resulting from the marginal improvement
in the statistics needed for evidence-based environmental policymaking.
Nevertheless, the collective improvement and modernisation of statistics
envisaged is likely to have a slight indirect positive impact by improving
the quality and availability of the relevant data of specific issues of
interest to environmental evidence-based policy-making.
The measures under policy option B.1 are not anticipated to drive any
direct environmental impacts.
Coherence with
overarching objectives
of EU policy
(-4 to +4)
Coherence 1 Relative to the baseline, policy option B.1 represents a marginal
improvement in the coherence of the ESOP framework with the
overarching objectives of EU policies. The ten measures included under
this option (3.2, 4.2, 6.2, 8.2, 9.2, 10.2, 11.2, 15.2, 17.2, 17.3) are
expected to bring about only a relatively small improvement in the
coherence of the ESOP framework compared to the measures proposed
under subsequent policy options.
Although a slight improvement over the baseline, the measures included
under this option may be insufficient to meet forthcoming EU policy
needs. For example, policymakers and NGOs require better data on
individuals and groups at risk of inequality and discrimination. This lack
of data has been highlighted as problematic, for instance, in the EU
Gender Equality Strategy 2020-2025, the LGBTIQ Equality Strategy, the
EU Roma Strategic Framework and the EU Anti-racism Action Plan
2020-2025. Policy option B.1 therefore scores lower than B.2, C.2, D.1
and D.2 which all include targeted measures (under measures 12-14)
aimed at improving the topical coverage of EU population statistics.
Efficiency
Incremental costs over the baseline
with range +/- 20%, in thousands of
2021 EUR
Eurostat MS/NSIs
One-off (over 3y) Recurrent (over
10 years)
One-off (over 3y) Recurrent (over
10 years)
96
148 - 222 74 – 110 4,420 – 6,630 5,294 – 7,940
Policy option B.1 entails a limited upgrade of statistical outputs and as
such would generate relatively minor additional costs to Member States
and Eurostat/the Commission, associated with developing, implementing,
monitoring and enforcing compliance with these more complex and
detailed outputs. The additional costs to Eurostat are mainly driven by
minor regulatory and enforcement costs associated with the new
measures, as well as one-off IT costs. For MS/NSIs, additional costs are
driven by one-off IT investment and administrative costs as well as
recurrent costs associated with enforcement, compliance and
administration of the new requirements and procedures.
It would also preserve the benefits generated by the current legal
framework. Additional benefits would be generated for Member States,
the Commission and data users in line with further improvements in the
quality, detail and understanding of European statistics on population, as
well as the quality and accuracy of evidence and media reporting. This
would likely lead to improvements in Member States’ ability to make
decisions, as well as the effectiveness/accuracy of policy at a local and
EU level, with associated economic and social benefits for citizens,
Member States and the EU as a whole. Eurostat would benefit from
improvements in its ability to meet changing policy needs and
reputational gains associated with improvements in the quality, detail and
quantity of data published. Administrative burdens may also be reduced
for data users, including non-institutional data users, in line with the
increased availability of data from a single reliable source (Eurostat).
Finally, a more flexible framework should generate efficiency gains in
administrative burden to adapt to evolving needs or new sources.
Overall assessment
Under policy option B.1, there will be limited changes to the status quo, including limited harmonisation,
separate statistical processes and limited update of statistical outputs.
In terms of its compliance with the proportionality principle, policy option B.1 is similar to the baseline (A)
and does not go beyond what is necessary to satisfactorily achieve the objectives.
This option would not be effective in addressing the problems identified with the current EU legal framework.
Differences in who is counted in European population figures will persist and statistical processes will remain
fragmented as in the baseline. While policy option B.1 makes improvements in terms of the ten measures
proposed, it does not make changes remarkable enough to be assigned a high comparative score, given the
broader effectiveness of more ambitious options in achieving the four specific objectives.
Regarding coherence with other EU policy objectives, policy option B.1 is expected to bring about relatively
minor positive impacts. Slightly positive impacts are expected on fundamental rights, as some of the measures
will indirectly contribute to the quality and availability of socio-demographic characteristics of specific
vulnerable and at-risk groups. Similarly, the measures under policy option B.1 are expected to bring about
relatively minor positive regulatory, environmental, and social impacts.
Concerning efficiency, this option is associated with relatively minor additional costs while upgrading the
statistical outputs produced. A limited upgrade may minimise the additional burden imposed while generating
added benefits by increasing, improving and rendering timelier the available European population data
available from Eurostat.
97
In summary, although a marginal improvement from the baseline, policy option B.1 would be suboptimal and
ineffective in addressing the problems identified in the current legislation. Both NSIs and non-NSIs identified
policy option B.1 as realistic (i.e. feasible to implement) in the validation workshops.
98
14.2.3. 1.2.3. Policy option B.2 - Limited harmonisation, separate statistical
processes, more expansive upgrade of statistical outputs and flexibility
Assessment criterion Score Brief description
Consistency with
proportionality
principle
(Pass /
Uncertain
/ Fall)
Does the option go
beyond what is
necessary to
satisfactorily achieve
the objectives?
Pass Option B.2 is similar to policy option B.1 in that it does not address
harmonisation of population definitions or statistical processes and entails
marginal changes compared to the baseline. The measures under this
option do not go beyond what is necessary to satisfactorily achieve the
objectives.
Is the scope of the
option limited to those
aspects that Member
States cannot achieve
satisfactorily on their
own, and where the
Union can do better?
Pass This option would only moderately extend statistical requirements
compared to the baseline, and some existing voluntary statistical topics
will become mandatory. The scope of the option is limited to those aspects
that Member States cannot achieve satisfactorily on their own, and where
the Union can do better (e.g. updated statistical requirements and
increased timeliness).
Are costs for the
Union, national
governments, regional
or local authorities,
economic operators or
citizens commensurate
with the objectives of
the initiative?
Pass The expected costs for the Union, national governments, regional or local
authorities, economic operators or citizens, are expected to increase
slightly, although they will remain minimal in relation to the objectives to
be achieved.
Is the form of action
(choice or instrument)
as simple as possible
and coherent with
satisfactory
achievement of the
objective and effective
enforcement?
Pass The updated statistical requirements will be introduced in demography,
migration, and census statistics to meet user needs on statistical topics
covered by the current EU legislation. The form of action (choice of
instrument) is thus as simple as possible and coherent with satisfactory
achievement of the objective and effective enforcement.
Effectiveness in
achieving the policy
objectives:
(-4 to +4)
SO1:
To ensure complete,
coherent and
comparable European
population statistics
1.5 Policy option B.2 is expected to have only a small positive impact on
SO1. The completeness of statistical outputs would be slightly improved
relative to the baseline (mostly as some voluntary statistics would become
mandatory under measures 12.2 and 13.2). However, this option would
not help to harmonise the population definitions or statistical processes
used by Member States. Option B.2 therefore scores slightly better than
Option B.1, but less well than Options C.1 to D.2.
SO2: 2 Policy option B.2 (like policy option C.2) performs better in ensuring the
availability of timely and frequent population statistics than policy options
99
To ensure the
availability of timely
and frequent
population statistics to
meet users’ needs
B.1 and C.1. Legal deadlines for census outputs are shortened (measure
9.2) and in topic 8 on timeliness, 8.3 ‘timeliness first’ is chosen over 8.2
‘quality first.’ Moreover, measure7.3 introduces infra-annual (quarterly)
data.
Policy option B.2 (as well as policy option C.2, D.1, and D.2) makes geo-
referenced data mandatory annually.
Policy option B.2 performs better in meeting SO2 than policy option B.1
(largely by introducing annual georeferenced data in measure 6.3 and
quarterly data in measure 7.3, and by prioritising timeliness in measure
8.3), but scores less well than policy options D.1 to D.2 (which go further
in providing annual LAU data in measure 5.2, detailed monthly data in
measure 7.2 and shorter census deadlines in measure 9.3). Overall,
policy option B.2 is expected to have a moderate positive impact on the
timeliness and frequency of population statistics.
SO3:
To provide statistics
that are sufficiently
comprehensive in
terms of relevant
topics and sufficiently
detailed in terms of
characteristics and
breakdowns
3 In policy option B.2 (similar to policy options C.2, D.1, and D.2), data on
current statistical output are significantly more detailed than the baseline
or either policy option B.1 or C.1 (more characteristics on migration
statistics through measures 10.3 and 11.3; improved disaggregation of
annual population data and migration statistics through measures 3.4 and
4.3).
In terms of comprehensiveness of relevant topics, policy option B.2 brings
new statistical topics including on equality and non-discrimination
(measure 14.2), and voluntary datasets and breakdowns are made
mandatory.
Overall, policy option B.2 (like policy options C.2, D.1 and D.2) scores
significantly better than the baseline and either policy option B.1 or C.1. It
is thus expected to have a strong positive impact against SO3.
SO4:
To promote
frameworks that are
sufficiently flexible to
adapt to evolving
policy needs and
opportunities from
new sources
2 Policy option B.2 (similar to policy option C.2) brings flexibility to the
overall framework and a legal mechanism for emerging needs is put in
place. Policy option B.2 (like policy option C.2) performs better than B.1
and C.1 due to introducing basic interoperability of statistical population
registers.
Overall, policy option B.2 is expected to have a moderate positive impact
in terms of SO4. By introducing a limited flexibility mechanism, this
option also reduces the risk of decreasing relevance of population statistics
compared to B.1 and C.1.
Impacts on
fundamental rights
1 Policy option B.2 represents a substantial improvement compared to the
baseline and policy option B.1 in respect of fundamental rights, as it
introduces equality data under measure 14.2. The other 13 measures under
this option would collectively indirectly improve fundamental rights, as
the measures are aimed at timeliness, completeness and more detail of
population statistics. For example, measure 4.3 aims for NSIs to provide
migration data at NUTS 3 level and measure 10.3 aims to define migrant
groups in more detail, including socioeconomic details. These measures
would facilitate evidence-based policy-making in respect of vulnerable
and at-risk groups at local administrative level.
100
However, it also poses a potential risk to fundamental rights (data privacy)
if sensitive data are collected and processed. Despite all necessary
safeguards112
, the feedback from consultees indicates that there is a trade-
off between increasing the availability of data that can be used to protect
groups at risk of inequality and discrimination, and the need to protect
citizens’ personal data, especially where sensitive topics are affected. This
is reflected with a score discount on net impact on fundamental rights.
Social impacts 2 Policy option B.2 represents an improvement over the baseline and policy
option B.1 in respect of social impacts, specifically the provision of social
statistics for informed evidence-based policy-making at EU and national
levels, by including statistics on migrant populations and flows (measures
10.3 and 11.3), on relevant societal topics currently only provided
voluntarily by the Member States (measures 12.2 and 12.3) and on the size
and demographic characteristics of particular groups (measure 14.2).
The collective improvement and modernisation of population statistics
envisaged is likely to have an indirect positive impact also by improving
the quality and availability of data for specific issues of interest to policy-
making, like equality or housing. Given the increased ambition in the level
of update of statistical outputs, associated benefits would be incrementally
higher than under policy options B.1 and C.1.
Environmental impacts2 Policy option B.2 represents an improvement compared to the baseline or
to policy option B.1 in respect of environmental impact, specifically the
annual grid data on housing (measure 6.3). Some of the measures (e.g. 3.2,
4.3 and 6.3) aim to improve the geographic disaggregation of the statistics,
which may indirectly benefit decision-making on environmental issues at
regional and local levels.
Given the increased ambition in the level of update of statistical outputs,
associated benefits would be incrementally higher than under policy
options B.1 and C.1.
This option is not anticipated to have any direct environmental impacts.
Overall, this option is expected to have a moderate (indirect) positive
environmental impact.
Coherence with
overarching objectives
of EU policy
(-4 to +4)
Coherence 2 The measures under policy option B.2 will bring about updated statistical
requirements in demography, migration, and population and housing
census statistics to better meet user needs on statistical topics covered by
the current EU legislation (e.g. supporting rural areas development, the
Green Deal, equality). A limited flexibility mechanism will allow the
framework to adapt to emerging needs, while basic interoperability among
Member States increases statistical coherence. The collective measures
112
For instance, the most sensitive data would be certain self-declared characteristics on grounds for discrimination
such as gender identity, race/ethnicity or sexual orientation. These would have to be collected through anonymised
voluntary surveys in which respondents cannot be in any way individually identified and that cannot be used for any
other purposes than statistics.
101
could be seen as a moderate improvement towards effective legal
coherence between the ESOP and new EU legal instruments.
Efficiency
Incremental costs over the
baseline with range +/- 20%, in
thousands of 2021 EUR
Eurostat MS/NSIs
One-off (over 3y) Recurrent (over
10 years)
One-off (over 3y) Recurrent (over
10 years)
294 – 440 282 - 422 13,506 – 20,260 25,822 – 38,732
Costs generated by policy option B.2 would be incrementally higher than
those in policy option B.1, as it entails a more ambitious upgrade of the
statistical outputs required of Member States. This assumes that most
processes/equipment required to deliver these requirements would be
more complex and therefore resource-intensive or costly. Additional
administrative, IT and compliance and enforcement costs will also be
incurred by MS/NSIs. For Eurostat, key cost drivers include more costly
one-off and recurring IT costs as well as additional
enforcement/monitoring costs associated with the additional requirements.
The inclusion of measures associated with gathering socio-economic data
on migrant populations as well as equalities data would also likely mean
all activities associated with data gathering are more costly. It is possible
that additional minor administrative burden would be imposed on citizens
in line with the requirement to respond to additional surveys (e.g. on
equality, see section 8.3.1). However, the overall effect is unclear, given
the potential for alignment/streamlining with other data-gathering
processes.
Benefits generated would be incrementally higher than those under policy
option B.1, including benefits from increased access to higher quality,
more detailed and timely data, as well as the associated improvements in
decision-making and policy-making. NSIs and Eurostat may benefit from
further increases in staff skills related to the production of population
statistics and related reputational gains. Additional benefits would be
generated, for example in access to accurate equality and non-
discrimination data (leading to associated economic and social benefits)
and an updated time series (leading to improvements in the quality and
efficiency of research studies). Eurostat (and to some extent NSIs) might
draw significant cost savings through reductions in the administrative
burden relating to coordination, review and quality assurance of voluntary
statistics (associated with relatively limited additional costs to NSIs and
Member States, as many already provide data voluntarily), as well as
adaptation to changing policy needs.
Overall, the expected efficiency is higher than for policy option B.1, as the
additional benefits generated are expected to add value (based on
stakeholder feedback, see Annex 2).
Overall assessment
Under policy option B.2, limited harmonisation and separate statistical processes will remain and an upgrade
of statistical outputs and more flexibility is proposed.
102
It is coherent with the proportionality principle. It does not go beyond what is necessary to satisfactorily
achieve the objectives, as the scope is limited to those aspects that Member States cannot achieve
satisfactorily on their own, and where the Union can do better (e.g. updated statistical requirements and
increased timeliness). The 14 measures would moderately extend statistical requirements and some existing
voluntary statistical topics would become mandatory. Differences in who is counted in European population
figures will persist and statistical processes will remain fragmented, as in the baseline.
The measures under policy option B.2 are expected to bring about a moderate positive impact on the overall
coherence of the ESOP framework with wider EU legislation, but with a somewhat limited harmonisation
and extent of changes compared to other options.
Positive impacts on fundamental rights are expected, as equality data are introduced and some other
measures will add socio-demographic characteristics of specific vulnerable and at-risk groups, while some
NSIs also raised concerns about data protection despite all necessary safeguards that would be put in place
(footnote 112). Similarly, the measures under policy option B.2 are expected to bring about moderate
positive environmental and social impacts.
Looking at efficiency, based on the stakeholder feedback, the measures included in policy option B.2 are
preferable to those under B.1 and could generate added value, although incrementally higher costs would
also be incurred. The feedback from both option validation workshops was mixed: A majority of NSIs stated
that the measures under policy Option B.2 would be unrealistic, while the opposite view was broadly
expressed during the non-NSI validation workshop.
103
14.2.4. 1.2.4. Policy option C.1 – Improved harmonisation, separate statistical
processes, limited update of statistical outputs
Assessment criterion Score Brief description
Consistency with
proportionality principle
(Pass /
Uncertain
/ Fall)
Does the option go
beyond what is
necessary to
satisfactorily achieve the
objectives?
Pass Although policy option C.1 goes further than policy options B.1 or B.2 in
addressing harmonisation of population definitions, statistical processes
and flexibility of the ESOP framework, it does not go beyond what is
necessary to satisfactorily achieve the objectives.
Is the scope of the option
limited to those aspects
that Member States
cannot achieve
satisfactorily on their
own, and where the
Union can do better?
Pass Policy option C.1 would only moderately extend statistical requirements
compared to the baseline, and some existing voluntary statistical topics
will become mandatory. Its scope is limited to those aspects that Member
States cannot achieve satisfactorily on their own, and where the Union
can do better (e.g. extending the range of statistical topics required and
improved harmonisation).
Are costs for the Union,
national governments,
regional or local
authorities, economic
operators or citizens
commensurate with the
objectives of the
initiative?
Pass The expected costs for the Union, national governments, regional or local
authorities, economic operators or citizens, are expected to increase
slightly, although remain minimal in relation to the objectives to be
achieved.
Is the form of action
(choice or instrument) as
simple as possible and
coherent with
satisfactory achievement
of the objective and
effective enforcement?
Pass The updated statistical requirements will be introduced in demography,
migration, and population and housing census statistics to meet user needs
on statistical topics covered by the current EU legislation. The form of
action (choice of instrument) is as simple as possible and coherent with
satisfactory achievement of the objective and effective enforcement.
Effectiveness in
achieving the policy
objectives:
(-4 to +4)
SO1:
To ensure complete,
coherent and comparable
European population
statistics
2 Policy option C.1 introduces a harmonised population base and otherwise
performs the same as policy option B.1. The harmonisation will lead to
substantially more coherent and comparable EU level data, therefore this
option performs better on SO1 than B.1 and also B.2 and would create
overall moderate positive impacts on SO1.
This also reduces the risk that this option does not go far enough in
improving the comparability of European population statistics.
SO2: 1 The single measure distinguishing this policy option from B.1 (measure
2.2. ‘harmonisation of the population definition’) is not expected to
104
To ensure the
availability of timely and
frequent population
statistics to meet users’
needs
improve achieving SO2, thus this option is ranked as equal to policy
option B.1 (see part 1.2.2 of this Annex).
SO3:
To provide statistics that
are sufficiently
comprehensive in terms
of relevant topics and
sufficiently detailed in
terms of characteristics
and breakdowns
1 The single measure distinguishing this policy option from B.1 (measure
2.2. ‘harmonisation of the population definition’) is not expected to
improve achieving SO3, thus this option is ranked as equal to policy
option B.1 (see part 1.2.2 of this Annex).
SO4:
To promote frameworks
that are sufficiently
flexible to adapt to
evolving policy needs
and opportunities from
new sources
1 The single measure distinguishing this policy option from B.1 (measure
2.2. ‘harmonisation of the population definition’) is not expected to
improve achieving SO4, thus this option is ranked as equal to policy
option B.1 (see part 1.2.2 of this Annex).
Impacts on fundamental
rights
1.5 Policy option C.1 introduces a harmonised population base and otherwise
performs the same as policy option B.1. The harmonisation will lead to
improved accuracy of the demographic representation of the EU
population in qualified majority voting of the Council, therefore this
option performs better on fundamental rights impact than B.1, but less
well than C.2 and D.1 to D.2.
Social impacts 1 The single measure distinguishing this policy option from B.1 (measure
2.2. ‘harmonisation of the population definition’) is not expected to have
any social impact, thus this option is ranked as equal to policy option B.1
(see part 1.2.2 of this Annex).
Environmental impacts 1 The single measure distinguishing this policy option from B.1 (measure
2.2. ‘harmonisation of the population definition’) is not expected to have
any social impact, thus this option is ranked as equal to policy option B.1
(see part 1.2.2 of this Annex).
Coherence with
overarching objectives
of EU policy
(-4 to +4)
Coherence 2 Policy option C.1 introduces a harmonised population base and otherwise
performs the same as policy option B.1. The harmonisation will lead to
improved coherence and comparability of information on the EU
population, therefore this option performs better on EU coherence than
B.1.
On the other hand, C.1 scores the same as B.2 overall on coherence,
because the lack of harmonisation in B.2 is balanced by the more
comprehensive data detail addressing EU policy priorities (rural
105
development, non, discrimination) and by a more powerful flexibility
mechanism.
Efficiency
Incremental costs over the baseline
with range +/- 20%, in thousands of
2021 EUR
Eurostat MS/NSIs
One-off (over 3y) Recurrent (over
10 years)
One-off (over 3y) Recurrent (over
10 years)
222 – 334 466 – 698 9,469 – 14,203 19,729 - 29,593
The costs and benefits generated by policy option C.1 mirror those of
policy option B.1, aside from the costs and benefits associated with
increased harmonisation of population definitions. Several Member States
would have to switch to a strict usual residence definition and therefore
incur additional costs, particularly one-off IT costs and costs associated
with the provision of training/guidance as well as recurring compliance,
enforcement and administrative costs. Eurostat would also incur minor
additional regulatory and enforcement costs.
Policy option C.1 would generate additional benefits compared to policy
option B.1, in line with the improved comparability across Member
States. This might improve the accuracy and efficiency of research
conducted and thus policy decisions based on population data. However,
harmonisation will not be achieved for all datasets and challenges in
respect of comparability of data will persist. Overall, it remains unclear if
the incremental benefits of improved harmonisation would outweigh the
incremental costs compared to B.1, and thus if C.1 would be considered
more efficient than B.1 or not.
Nonetheless, the public consultation suggested that allowing justified
exceptions in harmonisation may be preferable, despite its limiting effects
on achieving the benefits of full harmonisation.
Overall assessment
Policy option C.1 would see improved harmonisation, separate statistical processes, update of statistical
outputs and some flexibility improvements.
It is coherent with the proportionality principle. Although policy option C.1 goes further than policy options
B.1 or B.2 in addressing harmonisation of population definitions, statistical processes and flexibility of the
legal framework, it does not go beyond what is necessary to satisfactorily achieve the objectives, as the scope
is limited to those aspects that Member States cannot achieve satisfactorily on their own, and where the Union
can do better (e.g. extending the range of statistical topics required and improved harmonisation). Policy
option C.1 fails to adequately improve on the baseline on statistical outputs regarding frequency and detail,
which has a high impact on overall effectiveness in achieving the specific objectives.
The measures under policy option C.1 are expected to have a moderate positive impact on the overall
coherence between the ESOP framework and wider EU legislation, with improved harmonisation of the
processes involved in the collection of population statistics overall.
Positive impacts on fundamental rights are expected, as some of the measures will indirectly contribute to the
quality and availability of socio-demographic characteristics of specific vulnerable and at-risk groups.
Similarly, the improved harmonisation under policy option C.1 are expected to bring about relatively minor
106
positive indirect regulatory, environmental and social impacts.
Concerning efficiency, the expected costs and benefits generated by the implementation of policy option C.1
mirror those of policy option B.1, aside from costs and benefits associated with the harmonisation of
population definitions. Feedback from NSIs, data users and the Commission’s Statistical Correspondents
indicates that policy option C.1 is feasible to implement.
107
14.2.5. 1.2.5. Policy option C.2 – Improved harmonisation, separate statistical
processes, more expansive upgrade of statistical outputs and flexibility
Assessment criterion Score Brief description
Consistency with
proportionality
principle
(Pass /
Uncertain
/ Fall)
Does the option go
beyond what is
necessary to
satisfactorily achieve
the objectives?
Pass Although policy option C.2 goes further than policy options B.1 or B.2 in
addressing harmonisation of population definitions, statistical processes
and flexibility of the ESOP framework, it does not go beyond what is
necessary to satisfactorily achieve the objectives.
Is the scope of the
option limited to those
aspects that Member
States cannot achieve
satisfactorily on their
own, and where the
Union can do better?
Pass Policy option C.2 would only moderately extend statistical requirements,
and some existing voluntary statistical topics will become mandatory. The
scope is limited to those aspects that Member States cannot satisfactorily
achieve on their own, and where the Union can do better (e.g. extending
the range of statistical topics required and improved harmonisation).
Are costs for the
Union, national
governments, regional
or local authorities,
economic operators or
citizens commensurate
with the objectives of
the initiative?
Pass The expected costs for the Union, national governments, regional or local
authorities, economic operators or citizens, are expected to increase
slightly, although remain minimal in relation to the objectives to be
achieved.
Is the form of action
(choice or instrument)
as simple as possible
and coherent with
satisfactory
achievement of the
objective and effective
enforcement?
Pass The updated statistical requirements will be introduced in demography,
migration, and population and housing census statistics to meet user needs
on statistical topics covered by EU legislation. The form of action (choice
of instrument) is as simple as possible and coherent with satisfactory
achievement of the objective and effective enforcement.
Effectiveness in
achieving the policy
objectives:
(-4 to +4)
SO1:
To ensure complete,
coherent and
comparable European
population statistics
2.5 Policy option C.2 introduces a harmonised population base and otherwise
performs the same as policy option B.2, including as regards its
prevalence over C.1. The harmonisation will lead to substantially more
coherent and comparable EU level data, therefore this option performs
better on SO1 than B.2 and would create overall strong positive impacts
on SO1.
This also reduces the risk that this option does not go far enough in
108
improving the comparability of European population statistics.
SO2:
To ensure the
availability of timely
and frequent
population statistics to
meet users’ needs
2 The single measure distinguishing this policy option from B.2 (measure
2.2. ‘harmonisation of the population definition’) is not expected to
improve achieving SO2, thus this option is ranked as equal to policy
option B.2 (see part 1.2.3 of this Annex).
SO3:
To provide statistics
that are sufficiently
comprehensive in
terms of relevant
topics and sufficiently
detailed in terms of
characteristics and
breakdowns
3 The single measure distinguishing this policy option from B.2 (measure
2.2. ‘harmonisation of the population definition’) is not expected to
improve achieving SO3, thus this option is ranked as equal to policy
option B.2 (see part 1.2.3 of this Annex).
SO4:
To promote
frameworks that are
sufficiently flexible to
adapt to evolving
policy needs and
opportunities from
new sources
2 The single measure distinguishing this policy option from B.2 (measure
2.2. ‘harmonisation of the population definition’) is not expected to
improve achieving SO4, thus this option is ranked as equal to policy
option B.2 (see part 1.2.3 of this Annex).
Impacts on
fundamental rights
2 The single measure distinguishing this policy option from B.2 (2.2.
‘harmonisation of the population definition’) is expected to have
beneficial impact on fundamental rights due to more accurate
demographic representation in qualified majority voting of the council.
Therefore, C.2 scores higher than B.2, but less well than D.1 to D.2.
Social impacts 2 The single measure distinguishing this policy option from B.2 (measure
2.2. ‘harmonisation of the population definition’) is not expected to affect
social impacts, thus this option is ranked as equal to policy option B.2 (see
part 1.2.3 of this Annex).
Environmental impacts2 The single measure distinguishing this policy option from B.2 (measure
2.2. ‘harmonisation of the population definition’) is not expected to affect
environmental impacts, thus this option is ranked equal to policy option
B.2 (see part 1.2.3 of this Annex).
Coherence with
overarching objectives
of EU policy
(-4 to +4)
Coherence 3 Policy option C.2 combines the benefits of an expansive upgrade of
statistical outputs and flexibility from B.2 with an improved
harmonisation of the population base firstly introduced in C.1 and thus
ranks better than B.2 and C.1 with an overall strong positive impact on
109
coherence.
Efficiency
Incremental costs over the
baseline with range +/- 20%, in
thousands of 2021 EUR
Eurostat MS/NSIs
One-off (over 3y) Recurrent (over
10 years)
One-off (over 3y) Recurrent (over
10 years)
368 – 552 673 – 1,009 18,555 – 27,833 40,257 – 60,385
Similar to policy option B.1 in relation to C.1, the costs and benefits
generated by policy option C.2 are expected to be the same as those of
policy option B.2, aside from the costs and benefits associated with
increased harmonisation of population definitions. Measure 2.3 (on
harmonisation) would likely lead to additional costs for Eurostat and NSIs
(specifically one-off IT costs and costs associated with the provision of
training/guidance as well as recurring compliance, enforcement and
administrative costs for MS/NSIs and minor additional regulatory and
enforcement costs for Eurostat).
Applying the same logic again here, it remains unclear if the incremental
benefits of improved harmonisation would outweigh the incremental costs
compared to B.2, and thus if C.2 would be considered more efficient than
B.2 or not.
Overall assessment
Policy option C.2 would see improved harmonisation, separate statistical processes, upgrade of statistical
outputs and improved flexibility. Differences in who is counted in European population figures will be
partially reduced and statistical topics covered by EU legislation will be extended in line with users’ needs.
Policy option C.2 is coherent with the proportionality principle and a ‘Pass’ score is justified.
The relatively strong performance of policy option C.2 in achieving the specific objectives is compounded
by evidence from the stakeholder consultations. The proposal for a flexible harmonised population definition
attracted strong support across all user groups, meaning that despite retaining the status quo on population
registers, policy option C.2 performs well on SO1. It also performs strongly on SO2, SO3 and SO4, thus can
be highly scored overall in achieving the specific objectives.
Positive impacts on fundamental rights are expected, as equality data are introduced and some other
measures will add socio-demographic characteristics of specific vulnerable and at-risk groups, while some
NSIs also raised concerns about data protection despite all necessary safeguards that would be put in place
(footnote 112). Similarly, the measures under policy option C.2 are expected to bring about relatively
positive indirect regulatory, environmental and social impacts.
Looking at efficiency, the costs and benefits generated by implementing policy option C.2 are expected to be
the same as those generated by policy option B.2, aside from costs and benefits associated with increased
harmonisation of population definitions. The net effect on efficiency compared to B.2 remains unclear.
Although this option scores relatively well in terms of effectiveness (surpassed only by Options D.1 and
D.2), a large majority of NSIs indicated that policy option C.2 is not realistic due to heavily increased burden
and resource needs. In contrast, non-NSIs and the Commission’s Statistical Correspondents indicated during
their respective workshops that policy option C.2 is realistic.
110
14.2.6. 1.2.6. Policy option D.1 – Full harmonisation, separate statistical process
(no requirement for statistical population registers), major upgrade of
statistical outputs, and flexible framework
Assessment
criterion
Score Brief description
Consistency with
proportionality
principle
(Pass /
Uncertain /
Fall)
Does the option go
beyond what is
necessary to achieve
the objectives
satisfactorily?
Pass Policy option D.1 goes further than all previous policy options in how it
addresses harmonisation of population definitions, statistical processes
and flexibility of the legal framework. However, this policy option does
not go beyond what is necessary to achieve the objectives satisfactorily.
Is the scope of the
option limited to
those aspects that
Member States
cannot achieve
satisfactorily on their
own, and where the
Union can do better?
Uncertain This option would significantly extend statistical requirements when
compared to the baseline, and all existing voluntary statistical topics will
become mandatory. The scope of the option might not be limited to those
aspects that Member States cannot achieve satisfactorily on their own, and
where the Union can do better (extending the range of statistical topics
required and full harmonisation).
Are costs for the
Union, national
governments,
regional or local
authorities, economic
operators or citizens
commensurate with
the objectives of the
initiative?
Pass The expected costs for the Union, national governments, regional or local
authorities, economic operators or citizens, are expected to slightly
increase although remain minimal in relation to the objectives to be
achieved.
Is the form of action
(choice or
instrument) as simple
as possible and
coherent with
satisfactory
achievement of the
objective and
effective
enforcement?
Uncertain The updated statistical requirements will be introduced in a new, single
framework, and statistical processes will be redeveloped, therefore the
form of action (choice of instrument) is not as simple as possible, and
coherent with satisfactory achievement of the objective and effective
enforcement, as the introduction of an entirely new integrated legal
instrument can be expected to entail additional adjustment efforts for some
of the stakeholders involved.
Effectiveness in
achieving the policy
objectives:
(-4 to +4)
SO1:
To ensure complete,
coherent and
3 Compared to policy option C.2, policy options D.1 and D.2 introduce full
harmonisation of population definitions (based on the strict usual
residence concept, measure 2.2). Moreover, all voluntary data are
regulated under measures 12.2, 13.2 and the time series is fully updated
111
comparable European
population statistics
(measure 16.3).
However, going the maximum ambition in terms of SO1 without
similarly strong facilitation measures increases the risk that many
Member States would not be able to fulfil these new requirements. This
could lead to excessive use of derogations or even dragging compliance
issues. This is accounted for with a small score discount on SO1.
Despite these risks, policy option D.1 would still entail strong positive
impacts under SO1, which justifies the second highest score on achieving
SO1, falling short only of the added value from statistical population
registers included only in option D.2.
SO2:
To ensure the
availability of timely
and frequent
population statistics
to meet users’ needs
3 Policy options D.1 and D.2 perform the best in ensuring the availability of
timely and frequent population statistics. Legal deadlines for census
outputs are shortened (measure 9.2) and in topic 8 on timeliness, 8.3
‘timeliness first’ is chosen over 8.2 ‘quality first.’
Policy option D.1 (as well as policy option B.2, C.2, and D.2) also makes
georeferenced data mandatory annually and measure 7.2 introduces
population statistics on a quarterly basis. Policy option D.1 (similar to
policy option D.2) also introduces population statistics at LAU level
annually.
Nevertheless, the risk discount discussed under SO1 also applies to SO2.
Collectively, the measures under policy option D.1 would still deliver a
strong positive impact on the timeliness and frequency of population
statistics, surpassed only by Option D.2.
SO3:
To provide statistics
that are sufficiently
comprehensive in
terms of relevant
topics and
sufficiently details in
terms of
characteristics and
breakdowns
3.5 In policy option D.1 (and similarly to policy option B.2, C.2, and D.2),
data on current statistical output are more detailed compared to the
baseline and compared to both policy options B.1 and C.1 (more
characteristics regarding migration statistics under measures 10.3 and
11.3, and improved disaggregation of annual population data and
migration statistics under measures 3.2, 3.4, 4.3 and 6.3).
In terms of comprehensiveness of relevant topics, policy option D.1 brings
new statistical topics including extensive equality data (measure 14.3),
and voluntary datasets and breakdowns are made mandatory (measures
12.2, 13.2).
Nevertheless, the risk discount discussed under SO1 also applies to SO3.
Overall, policy option D.1 introduces strong positive impacts under SO3
and thus scores better than the baseline and policy options B.1, B.2, C.1,
and C.2, falling slightly short of D.2 only due to the risk discount.
SO4:
To promote
frameworks that are
sufficiently flexible
to adapt to evolving
policy needs and
opportunities from
3.5 Policy option D.1 (like policy option D.2) offers highly flexible legal
frameworks around data collection and seeks mechanisms to adapt
statistics more quickly and efficiently (e.g., to meet emerging user needs,
exploit new data sources or methods). Moreover, strong enablers on
improved use of administrative and new sources as well as full EU level
interoperability of existing statistical population registers are introduced.
112
new sources
Impacts on
fundamental rights
3 Policy option D.1 will bring about full harmonisation and differences in
who is counted in European population figures will be removed by the
introduction of a single, harmonised population definition. New statistical
topics, particularly regarding equality, will cover comprehensively present
and emerging user needs. This will contribute to improving the quality and
availability of the relevant data of specific vulnerable and at-risk groups
and lead to better informed policy making.
However, considering the significant ambition entailed in these measures,
risks presented under SO1 and SO3 apply here too. Moreover, as noted
previously, despite all necessary safeguards to be put in place (see
footnote 112) there is a trade-off between increasing the availability of
data that can be used to protect groups at risk of inequality and
discrimination, and the need to protect citizens’ anonymity, especially
where sensitive data are to be collected and processed.
Overall, including a risk discount policy option D.1 would introduce
strong positive impacts on fundamental rights.
Social impacts 3 Policy option D.1 represents an improvement in comparison to option C.2
concerning social impacts – and more specifically, provision of social
statistics for informed evidence-based policy making at EU and national
level. The measures under this option, and increased level of ambition of
the upgrade of statistical outputs across measures would collectively
improve social impacts since the measures are aimed at inter alia
timeliness, harmonisation and detail of population statistics collected.
Overall this option is expected to have strong positive social impacts.
However, considering the significant ambition entailed in these
measures, risks presented under SO3 apply here too.
Overall, including a risk discount policy option D.1 would introduce
strong positive social impacts.
Environmental
impacts
3 Policy option D.1 represents an improvement in comparison to the
baseline and option C.2 concerning environmental impact. The measures,
and in particular the increased level of ambition of the statistical output
associated with certain measures (particularly measure 6.3) under this
option would collectively improve environmental impacts.
However, considering the significant ambition entailed in these
measures, risks presented under SO3 apply here too.
Overall, Option D.1 is expected to have a strong positive environmental
impact.
Coherence with
overarching
objectives of EU
policy
(-4 to +4)
Coherence 4 The measures under policy option D.1 are expected to bring about mostly
positive impacts to the overall coherence of the ESOP framework within
113
the wider EU landscape by means of full harmonisation and a major
upgrade of statistical outputs in comparison to the measures proposed
under previous policy options. The measures under this policy option will
ensure that statistical topics will cover comprehensively present and
emerging user needs. Moreover, flexibility mechanisms will allow the
ESOP framework to easily adapt to emerging needs as well by allowing to
embrace new data sources as they emerge. Finally, full interoperability is
in line with wider aims for efficiency gains through digitalisation.
Therefore, the collective measures would create very strong positive
impacts towards an effective legal coherence between the ESOP and other
new EU legal instruments.
Efficiency
Incremental costs over the
baseline with range +/- 20%, in
thousands of 2021 EUR
Eurostat MS/NSIs
One-off (over 3y) Recurrent (over
10 years)
One-off (over 3y) Recurrent (over
10 years)
444 – 666 898 – 1,346 31,470 – 47,204 57,458 – 86,186
Option D.1 is significantly more ambitious than option C.2 regarding
statistical outputs, harmonisation and flexibility and as such generates
higher costs and benefits. All benefits generated in B.2 and C.2 would also
be incrementally higher in line with these improvements.
Costs in relation to the requirement for strict harmonisation, though
significant, would only be incurred by those Member States that do not
currently apply the usual residence definition. Member States would also
generate higher costs relative to options B.2 and C.2, in line with the
added complexity and requirements from the upgraded statistical output.
In particular, the added detail required within measures associated with
gathering socio-economic data on migrant populations as well as equality
data would also likely mean all activities associated with data gathering
are more costly activities. Therefore, costs to MS are driven by higher
costs in all cost categories relative to policy options B.2 and C.2.
Relative to baseline situation, Eurostat would incur costs associated with
the preparation and development of measures, IT investment to cope with
both the receipt of upgraded statistical outputs across all Member States.
Enforcement costs were estimated to be relatively minor in relation to the
harmonisation of population definitions, however. Additional burden are
likely to be incurred by citizens, in line with the increased burden
associated with responding to new surveys.
Benefits generated by this option are also likely higher than those in all
other options, given the increased level of ambition and number of
measures included. Achieving full harmonisation would likely lead to
improvements in the accuracy of policy and budgetary decisions regarding
the population (or reduced costs based on inaccuracies of those decisions)
and improved efficiency and accuracy of comparative research produced
across Member States. It might also reduce the administrative burden to
Eurostat associated with adjusting non-harmonised data as well as in
conducting checks on data provided voluntarily (as in B.2 and C.2).
114
Although it has not been possible to quantify the benefits of this option,
the available anecdotal evidence from the consultation seems to suggest
that it is unlikely that the benefits will outweigh the very high additional
costs associated with this option.
Overall assessment
Under policy option D.1, there will be full harmonisation of the population definition and improvements
made to the flexibility of the EU legal framework. A major upgrade of statistical outputs will also be
provided. However, unlike Option D.2, there will not be any requirement for Member States to adopt
statistical population registers. In terms of its compliance with the proportionality principle, policy option
D.1 goes further than other policy options in addressing the harmonisation of population definitions,
increasing statistical outputs and enhancing the flexibility of the legal framework. However, the expected
costs are expected to be much higher than for the previous policy options. Thus, this level of EU action may
not be proportionate to the objectives of the ESOP initiative and thus, it is uncertain whether this option fully
complies with the proportionality principle. Indeed, policy option D.1 was not viewed very favourably by
NSIs during the options validation workshop even by NSIs that already use administrative or statistical
population registers. In the workshop, NSIs voted unanimously that policy option D.1 is unrealistic due to
very heavily increased burden and resource needs, despite potential benefits expected for NSIs (see Table 6).
However, the Commission Statistical Correspondents and non-NSIs, in their respective validation
workshops, generally stated that the option is realistic.
Policy option D.1 is expected to have a very strong positive impact on coherence. In particular, the flexibility
to be provided under Option D.1 will help to ensure that European population statistics and the EU legal
framework governing their collection can be adapted, as needed, to align with users’ needs as well as new
data sources, and with wider EU policy and legislation in the years to come.
In terms of effectiveness, policy option D.1 scores second highest just below D.2. The theoretical
effectiveness under every specific objective is very high, and the results of the consultation tend to
corroborate this predicted effectiveness by returning strong endorsements from a variety of statistical users.
Very positive impacts are expected on fundamental rights as extensive equality data are introduced and the
other measures will indirectly contribute to the quality and availability of socio-demographic characteristics
of specific vulnerable and at-risk groups and consequently to a better informed policy making. However,
despite all necessary safeguards (see footnote 112) the benefits of improved data available are also offset to
some extent by concerns over data privacy. Similarly, the measures under policy option D.1 are also
expected to bring about positive indirect environmental and social impacts within improved harmonisation.
Concerning efficiency, the high costs and burdens associated with option D.1 especially during an extended
adjustment phase are clear, yet, despite associated benefits being present (see Table 6), the extent to which
the benefits outweigh these costs are less certain.
115
14.2.7. 1.2.7. Policy option D.2 – Full harmonisation, redeveloped and integrated
statistical processes in a single and flexible framework, major upgrade of
statistical outputs
Assessment
criterion
Score Brief description
Consistency with
proportionality
principle
(Pass /
Uncertain /
Fall)
Does the option go
beyond what is
necessary to
satisfactorily achieve
the objectives?
Pass Policy option D.2 goes further than any other policy options in addressing
harmonisation of population definitions, statistical processes and
flexibility of the ESOP framework. However, it does not go beyond what
is necessary to satisfactorily achieve the objectives.
Is the scope of the
option limited to
those aspects that
Member States
cannot achieve
satisfactorily on their
own, and where the
Union can do better?
Uncertain Policy option D.2 would significantly extend statistical requirements
compared to the baseline, with all existing voluntary statistical topics
becoming mandatory. The scope might not be limited to those aspects that
Member States cannot achieve satisfactorily on their own, and where the
Union can do better (extending the range of statistical topics required and
full harmonisation).
Are costs for the
Union, national
governments,
regional or local
authorities, economic
operators or citizens
commensurate with
the objectives of the
initiative?
Pass The expected costs for the Union, national governments, regional or local
authorities, economic operators or citizens, are expected to slightly
increase, although remain minimal in relation to the objectives to be
achieved.
Is the form of action
(choice or
instrument) as simple
as possible and
coherent with
satisfactory
achievement of the
objective and
effective
enforcement?
Uncertain Statistical processes will be redeveloped and integrated. The form of
action (choice of instrument) is not as simple as possible and coherent
with satisfactory achievement of the objective and effective enforcement,
as the introduction of an entirely new integrated legal instrument can be
expected to entail additional adjustment efforts for some stakeholders.
Effectiveness in
achieving the policy
objectives:
(-4 to +4)
SO1:
To ensure complete,
coherent and
4 Policy option D.2 (like policy option D.1) introduces the full
harmonisation of population definitions (based on a strict usual residence
concept). Going beyond policy option D.1, In policy option D.2, NSIs are
also required to implement national statistical registers which should
116
comparable European
population statistics
further improve coherence and comparability. Moreover, the EU wide
introduction of efficient integrated statistical production systems
ultimately reduces risks of compliance issues in the long term. Therefore,
policy option D.2 performs best on SO1, creating very strong positive
impacts.
SO2:
To ensure the
availability of timely
and frequent
population statistics
to meet users’ needs
4 The single measure distinguishing this policy option from D.1 (measure
2.2. ‘statistical population registers’) is expected to reduce the risks in
achieving SO2 compared to policy option D.1 (as for SO1 above), thus
this option will introduce very strong positive impacts on SO2 and thus
score higher than policy option D.1 (see part 1.2.6 of this Annex).
SO3:
To provide statistics
that are sufficiently
comprehensive in
terms of relevant
topics and
sufficiently detailed
in terms of
characteristics and
breakdowns
4 The single measure distinguishing this policy option from D.1 (measure
2.2. ‘statistical population registers’) is expected to reduce the risks in
achieving SO3 compared to policy option D.1 (as for SO1 above), thus
this option will introduce very strong positive impacts on SO3 and thus
score higher than policy option D.1 (see part 1.2.6 of this Annex).
SO4:
To promote
frameworks that are
sufficiently flexible
to adapt to evolving
policy needs and
opportunities from
new sources
4 The single measure distinguishing this policy option from D.1 (measure
2.2. ‘statistical population registers’) is expected to reduce the risks in
achieving SO4 compared to policy option D.1 (as for SO1 above), thus
this option will introduce very strong positive impacts on SO4 and thus
score higher than policy option D.1 (see part 1.2.6 of this Annex).
Impacts on
fundamental rights
3 The single measure distinguishing this policy option from D.1 (measure
2.2. ‘statistical population registers’) is expected to reduce the risks in
achieving the relevant measures (in particular 14.3) compared to policy
option D.1. On the other hand, according to NSI opinions, the introduction
of a statistical population register creates itself sizeable additional
fundamental rights concerns (voiced e.g. by NSIs in the stakeholder
consultation) in those countries that do not have such a system yet, mostly
around the proportionality of compiling and/or interlinking such amounts
of personal data even if solely for statistical purposes. This is accounted
for with a risk discount on this score.
Overall, this option will introduce very strong positive impacts on SO4
and thus score higher than policy option D.1 (see part 1.2.6 of this
Annex), but the risk discount is estimated to roughly balance this.
Social impacts 3 The single measure distinguishing this policy option from D.1 (measure
2.2. ‘statistical population registers’) is expected to reduce the risks in
achieving the relevant measures to improve evidence for policies on
labour market integration, social inclusion and equality (measures 3.2, 3.3,
117
6.3, 7.2, 11.3, 12.2, 14.3) compared to policy option D.1. On the other
hand, according to some NSI opinions, introducing a statistical population
register may likely be perceived as a very sensitive issue in some of the
Member States where no such system exists yet, especially when
combined with EU level interoperability elements. In the worst case, this
might trigger sizeable negative public attention that may question the
social license of producing official population statistics overall. This is
accounted for with a risk discount on this score.
Overall, this option will introduce very strong positive impacts on SO4
and thus score higher than policy option D.1 (see part 1.2.6 of this
Annex), but the risk discount is estimated to roughly balance this.
Environmental
impacts
4 The single measure distinguishing this policy option from D.1 (measure
2.2. ‘statistical population registers’) is expected to reduce the risks in
achieving the relevant measures (in particular 6.3) compared to policy
option D.1. Moreover, the introduction of statistical population registers
would greatly reduce the efforts compared to traditional censuses, and
thus minimise the environmental footprint of the census.
Overall, this option will introduce very strong (indirect) positive
environmental impacts and thus score higher than policy option D.1 (see
part 1.2.6 of this Annex).
Coherence with
overarching
objectives of EU
policy
(-4 to +4)
Coherence 4 The single measure distinguishing this policy option from D.1 (measure
2.2. ‘statistical population registers’) is not expected to improve coherence
with EU policy objectives beyond the goals already in option D.1, thus
this option is ranked as equal to policy option D.1 (see part 1.2.6 of this
Annex).
Efficiency
Incremental costs over the
baseline with range +/- 20%, in
thousands of 2021 EUR
Eurostat MS/NSIs
One-off (over
3y)
Recurrent (over
10 years)
One-off (over 3y) Recurrent (over
10 years)
629 – 943 1,391 – 2,087 38,323 – 57,485 97,981 – 146,971
Policy option D.2 is the most ambitious option and differs from D.1 only
in the added requirement to use statistical population registers.
While significant, costs related to the establishment of statistical
population registers would only be incurred by those Member States that
do not and are not already planning to develop a population register.
Eurostat would incur additional costs associated with the preparation and
development of this additional measure relative to D.1, as well as IT
investment to create the necessary infrastructure for a European
framework of statistical population registers including interoperability
118
mechanisms.
The use of population registers would likely lead to further improvements
in the accuracy of policy and budgetary decisions and improved efficiency
and accuracy of research. The use of population registers would lead to
benefits, including to Eurostat and the wider EU, through increased
flexibility and ease of access to up-to-date data, as well as to Member
States (those that do not or are not planning to implement a population
register) and citizens, for example, associated with reduced long-term
costs and administrative burdens associated with the production of data,
including more efficient delivery of censuses. Overall, all types of
statistics under this initiative, and many other social statistics products
beyond population statistics, are expected to profit in the form of higher
statistical quality at EU level (such as through better sampling frames for
social surveys based on samples).
As with measure D.1, however, despite these clear benefits, the extent to
which they outweigh the very high additional costs associated with this
option remains uncertain. Nevertheless, even in absence of a more robust
efficiency benchmark against the B – C options, the long-term efficiency
gains introduced by the statistical registers clearly lead to a relatively
better efficiency performance than D.1: All incremental costs and benefits
except entailed by the statistical registers are exactly the same, but the
sizeable upfront investment required in such statistical registers in the
affected Member States (group 3 in section 6.1) will very likely create
sustainable long-term efficiency gains in producing all kinds of ambitious
statistical outputs.
Overall assessment
Policy option D.2 differs from D.1 in that it mandates the use of statistical population registers. Therefore, it
goes further than the other policy options in addressing the coherence and flexibility of the ESOP
framework. It also goes further than all the other options in addressing the four specific objectives and is thus
scored as the most effective option overall However, the expected costs would be higher, and the mandatory
introduction of statistical population registers might not be simple or coherent with satisfactory achievement
of the objective and effective enforcement. It is therefore uncertain whether this option is coherent with the
proportionality principle. Indeed, despite the potentially significant long-term benefits expected, the NSIs
that participated in the validation workshop unanimously agreed that policy option D.2 is unrealistic, largely
due to the costs involved in transitioning to national statistical population registers. Conversely, this option
was largely viewed positively by data users. The measure on establishing statistical population registers was
viewed particularly positively by the data users who responded to the public consultation.
In terms of coherence, policy option D.2 is expected to have very strong positive impacts on the overall
ESOP framework relative to the baseline within the wider EU landscape by means of a full harmonisation
and redeveloped and integrated statistical processes. Such a framework would represent the state of the art in
flexibility to adapt as users’ needs change and as new sources of data and methods become available.
Positive environmental impacts are also expected, in line with likely simplifications of processes and
reductions in burden, as well as greening the data collection procedures by using statistical population
registers. Fundamental rights and social impacts are stronger than in option D.1 through the statistical
population registers, but this is balanced by specific risk concerns around data protection and a sufficient
social license for such endeavours.
Looking at efficiency, as with D.1 the high costs and burdens associated with policy option D.2 especially
during an extended adjustment phase are clear, as are the benefits, but the extent to which the benefits
119
outweigh the costs is less certain. The overall efficiency of this option is therefore judged qualitatively as
negative compared to other B – C options (that are less ambitious in delivering on the objectives but also
have more modest – and less uncertain – incremental cost estimates). Nevertheless, the long-term efficiency
gains introduced by the statistical registers lead to a better efficiency performance than D.1 (introducing
most ambitious statistical outputs and flexibility mechanisms without any action to foster efficiency
improvements).
Overall, if the subsidiarity and proportionality concerns around statistical population registers are considered
less tangible, option D.2 prevails overall due to very strong positive impacts despite question marks
remaining around its efficiency.
120
15. 2. SUMMARY OF COSTS AND BENEFITS
I. Overview of Benefits (total for all provisions) – Preferred Option
Description Amount Comments
Direct benefits
Better information on own local/regional environment Not quantified People (citizens and migrants)
Reduced response burden Not quantified
Better EU level timeliness and completeness of
statistics across all Member States
Not quantified EU level institutional users
Better EU level comparability and coherence of
statistics across all Member States
Not quantified
More accurate and comparable total population counts
for Council voting
Not quantified
Improved inputs to demographic change monitoring
and projecting the long-term budget sustainability in
relation to population ageing
Not quantified
Better data evidence for monitoring and policy-
making
Not quantified
Better comparability with other Member States and
EU regions
Not quantified Other institutional users (national and sub-
national levels)
Better data evidence (through grids) for policy-
making in border regions and local crisis response
Not quantified
Better research/analysis incl. improved comparability
across the EU
Not quantified Other professional users
Reduced administrative burden (through ability to
find all needed statistics on Eurostat website)
Not quantified
Economic benefits from better European statistics
available
Not quantified
Improved comparability of statistics with other
Member States
Not quantified Statistics producers (NSIs)
Improved accuracy and coverage of statistics due to
interoperability with other Member States
Not quantified
Reduced administrative burden (through simplified
statistics transmission processes)
Not quantified
Reduced administrative burden (through integrated
statistics production process)
e.g. between EUR 141
million and EUR 563
million per EU census
round
Reduced administrative burden (through improved
use of administrative and/or other data sources)
Not quantified
Reduced administrative burden (relating to regulatory
changes to adapt to evolving policy needs)
Not quantified
Reduced administrative burden through streamlined
data exchange with NSIs
Not quantified Administrative data providers
Increased value added from own data through
improved reuse
Not quantified
Improved legal base of statistical cooperation through
clear mandate
Not quantified
121
Better data evidence for other European statistics (e.g.
sample surveys, national accounts)
Not quantified Eurostat
Improved collaboration with ESS partners due to EU
wide interoperability
Not quantified
Reduced administrative burden (relating to regulatory
changes to adapt to evolving policy needs)
Not quantified
Reduced administrative burden (related to
coordination/ quality assurance for voluntary data)
Not quantified
Indirect benefits
Benefits from improved policy-making Not quantified People
Reputational gains from improved policy-making and
EU decision-making in general
Not quantified EU level institutional users
Reputational gains from improved visibility and
transparency in a European context
Not quantified Other institutional users (national and sub-
national levels)
Increased ability to meet legal requirements Not quantified Statistics producers (NSIs)
Increased staff skills Not quantified
Improvements in administrative registers thanks to
closer collaboration with NSI
Not quantified Administrative data providers
Reputational gains from improved reuse of
administrative registers
Not quantified
Advancement of Eurostat mission ‘to provide high
quality statistics and data on Europe’
Not quantified Eurostat
Improved collaboration with EU level policy users Not quantified
Reputational gains from enhanced international
standing of European statistics free for all
Not quantified
(1) Estimates are gross values relative to the baseline for the preferred option as a whole (i.e. the impact of individual
actions/obligations of the preferred option are aggregated together); (2) Please indicate which stakeholder group is the
main recipient of the benefit in the comment section;(3) For reductions in regulatory costs, please describe details as to
how the saving arises (e.g. reductions in adjustment costs, administrative costs, regulatory charges, enforcement costs,
etc.;); (4) Cost savings related to the ’one in, one out’ approach are detailed in Tool #58 and #59 of the ‘better
regulation’ toolbox. * if relevant
II. Overview of costs – Preferred option (in thousands of 2021 EUR with a +/- 20% uncertainty margin)
Citizens/Consumers Businesses Administrations
One-off Recurrent One-off Recurrent One-off Recurrent
Policy
option D.2
Direct adjustment
costs
- - - -
20,537 – 30,807 N/A
Direct
administrative
costs
- - - -
N/A 27,970 – 41,956
IT costs - - - - 20,464 – 30,679 14,569 – 21,854
Direct regulatory
fees and charges
- - - -
Not estimated Not estimated
Direct
enforcement costs
- - - -
N/A 62,062 – 92,793
Indirect costs - - - - Not estimated Not estimated
122
(1) Estimates (gross values) to be provided with respect to the baseline; (2) costs are provided for each identifiable
action/obligation of the preferred option otherwise for all retained options when no preferred option is specified; (3) If
relevant and available, please present information on costs according to the standard typology of costs (adjustment
costs, administrative costs, regulatory charges, enforcement costs, indirect costs;). (4) Administrative costs for
offsetting as explained in Tool #58 and #59 of the ‘better regulation’ toolbox. The total adjustment costs should equal
the sum of the adjustment costs presented in the upper part of the table (whenever they are quantifiable and/or can be
monetised). Measures taken with a view to compensate adjustment costs to the greatest extent possible are presented in
the section of the impact assessment report presenting the preferred option.
16. 3. RELEVANT SUSTAINABLE DEVELOPMENT GOALS
The preferred policy option D.2 covers various measures to improve the EU level data evidence relevant for several sustainable
development goals (SDGs). However, impacts in terms of actual progress towards any of the SDGs stemming from improved data
evidence are naturally indirect and hard to assess. For instance, if used for improved SDG policy-making, actual progress would
largely depend on dispersed contextual factors and on the sensitivity of policy impacts on the availability and quality of data
evidence. Therefore, the table below lists the relevant SDGs including improved data evidence expected under the relevant measures
included in the preferred option (cf. Figure 9 and
123
Table 2), but does not analyse the indirect policy impacts further, in line with the general approach
taken in section 6.2.
III. Overview of relevant Sustainable Development Goals – Preferred Option(s)
Relevant SDG Expected progress towards the
Goal
Comments – new/improved collection of
1 – No poverty Better data evidence for monitoring
and relevant policies
 annual housing data incl. homelessness (measure 6.3)
 socio-economic details of migrants (measures 10.3, 11.3)
 equality data incl. on vulnerable groups based e.g. on
race/ethnicity (measure 14.3)
3 – Good health and well-
being
 more frequent and detailed demographic incl. mortality
data for better indicators like regional or local life
expectancy (measures 3.2, 3.3, 5.2, 7.2, 12.2)
 socio-economic details of migrants (measures 10.3, 11.3)
4 – Quality education  mandatory annual data on education details of population
stocks (measure 13.2)
5 – Gender equality  equality data incl. on gender identity (measure 14.3)
9 – Industry, innovation
and infrastructure
 annual georeferenced housing data incl. energy efficiency
characteristics (measure 6.3)
10 – Reduced inequalities  equality data incl. on vulnerable groups (measure 14.3)
11 – Sustainable cities and
communities
 annual population data for cities and FUAs (measure 3.3)
 annual georeferenced population and housing data
(measure 6.3)
13 – Climate action  annual georeferenced housing data incl. energy efficiency
characteristics (measure 6.3)
15 – Life on land  annual population data by DEGURBA (measure 3.3)
 infra-annual population data (measure 7.2)
 annual georeferenced population and housing data
(measure 6.3)
 migration data at NUTS 3 regional level (measure 4.3)
16 – Peace, justice and
strong institutions
 equality data incl. on vulnerable groups based e.g. on
race/ethnicity (measure 14.3)
124
ANNEX 3: ANALYTICAL METHODS
The contract support study on this impact assessment has assessed a total of seven options, namely:
 A – Baseline scenario
 policy option B.1 – Limited harmonisation, separate statistical processes, limited upgrade of
statistical outputs
 policy option B.2 – Limited harmonisation, separate statistical processes, more expansive
upgrade of statistical outputs and flexibility
 policy option C.1 – Improved harmonisation, separate statistical processes, updated
statistical outputs
 policy option C.2 – Improved harmonisation, separate statistical processes, more expansive
upgrade of statistical outputs and flexibility
 policy option D.1 – Full harmonisation, separate statistical processes and flexible
framework, major upgrade of statistical outputs
 policy option D.2 – Full harmonisation, redeveloped and integrated statistical processes in a
single and flexible framework, major upgrade of statistical outputs
Costs and benefits have been considered for four main stakeholder groups, namely 1) Member
States and their National Statistical Institutes (NSIs), 2) the European Commission, including
Eurostat specifically 3) employers / businesses / non-institutional data users and 4) EU citizens and
third-country nationals. Whereas cost and benefit items were identified and considered for all four
groups, costs estimates could only be quantified for the first two and benefits were generally not
appropriate for quantification. This was in part due to a lack of available data, for example on the
costs to citizens of participating in census rounds. More generally this was due to their effects being
more ambiguous and variable across Member States and stakeholder groups. For example, the
benefits to non-institutional data users from increased access to high quality European statistics on
population would be challenging to quantify since this would depend on several additional factors,
such as how this data would be used or the cost of accessing data through alternative sources.
Therefore, estimates for benefits are not available. For the same reason, cost estimates should be
considered with caution: it is likely that these costs have been underestimated and represent a
minimal figure of overall costs incurred by the different groups of stakeholders.
17. 1. OVERALL APPROACH
Our overall approach to the estimation of costs and benefits consisted of the following key steps:
Firstly, the cost and benefit items associated with each policy measure were identified and itemised,
considering the type of cost (i.e., one-off or recurrent costs, overall cost categories), the stakeholder
group impacted and, in the case of Member States, the proportion and/the extent to which these
Member States were likely to be impacted by the different measures. This itemisation was reviewed
and refined in cooperation with Eurostat.
125
As noted above, it was determined, and agreed with Eurostat, that benefits were not possible to
quantify. These have however been qualitatively defined in the main report, as well as in more
details in Annex 2.
In consultation with Eurostat, it was agreed to assess the costs and benefits associated with each of
the seven policy options. The quantification of these costs has been carried out considering the
impact of each policy option as a whole rather than assessing the impact of each measure
individually to then aggregating these into options.
Overall, estimates and assumptions were based on a combination of several factors or criteria,
including:
 The implementation of statistical registers;
 The introduction of more harmonised population definition based on the usual residence
concept; and,
 The different level of ambitions of each option in terms of upgrade of statistical outputs
Costs and costs savings for each cost item and (group of) policy measures were then aggregated
across Member States, where relevant.
Finally, estimated costs and cost savings were aggregated for each policy option and estimated by
stakeholder group (Member States and the European Commission) and cost type (i.e. one-off and
recurrent. Details of what this aggregation process consisted of, as well as specific calculations and
assumptions applied to estimate costs for each stakeholder group and limitations of the model are
set out below.
18. 2. SOURCES
The assessment of the costs and the benefits was carried out using multiple sources and
triangulating data when possible. The main sources used have been:
1. Inputs provided by Eurostat, including through regular meetings/feedback requests as well
as data around administrative, grant, IT/infrastructure and contract-related costs to Eurostat
associated with the three in-scope regulations, a review of the completeness of voluntary and
mandatory statistics over time and a survey of Member States’ regarding costs associated
with population statistics.
2. Data gathered throughout the research study, including the workshops, literature review,
NSI survey and Open Public Consultation
3. The study team members’ experience of conducting similar quantification exercises, in
particular on the cost of reporting to the EU, training of staff, familiarisation with EU
legislation, transposition, and compliance costs. The approach is similar to one used most
recently for a DG HOME Study assessing the impacts of possible revisions to the Long
Term Residency and Single Permit Directives in 2021 (positive opinion of the RSB in
October 2021), DG JUST Study on the impacts of a possible revision of the Consumer
Credit Directive (CCD) in 2020-2021 (positive opinion of the RSB in May 2021), and DG
126
HOME Evaluation of the Counter-Terrorism Directive (positive opinion of the RSB in July
2021), among others in previous years.
19. 3. GENERAL ASSUMPTIONS
19.1. 3.1. Costs itemisation
The IA study assesses the impact of seven policy options, including the baseline, on different cost
types.
Costs likely to be borne by the Commission (Eurostat) were itemised as follow:
 Regulatory costs: Costs required to prepare and draft the new regulation, including
adopting implementing acts, communication, and provision of training, working group
meetings managing the overall delivery of the intervention. These are one-off costs,
occurring over a 3-year transition period
 Enforcement costs: Resources required to monitor MS alignment to, and compliance with
the requirements of the new regulation (including around data protection, data quality and
transparency) and to receive and publish data provided. These are recurrent costs, occurring
over the 10-year period
 IT costs: Costs associated with new software/application/IT update necessary for
implementing the new measure (one-off costs), as well as the maintenance of IT equipment
(recurrent costs)
The following cost items were identified as being the main costs likely to be borne by the Member
States and their respective NSIs:
 Administrative costs: these are the costs associated with any resources required to develop,
implement, and communicate any training required (one-off costs), as well as any costs
associated with the design, implementation and running of (new) procedures (recurrent
costs)
 Compliances costs: these are the costs with resources required to ensure compliance with
requirements, including adopting national arrangements or legislation to ensure cooperation
with relevant source data holders (recurrent costs)
 Enforcement costs: these are the costs associated with resources required to monitor and
ensure compliance to the regulation in the MS (recurrent costs)
 IT costs: Costs associated with new software/application/IT update necessary for
implementing the new measure (one-off costs), as well as the maintenance of IT equipment
(recurrent costs)
19.2. 3.2. Assessment period and social discount rate
The study considers the impact of the various policy options considered over a 10-year period from
Year 0 to Year 10.
127
For this impact assessment, it was assumed that all adjustment costs (or otherwise so-called “one-
off” costs) would occur over a 3-year transition period, from Year 0 to Year 2 included. All on-
going (also so-called “recurrent”) costs have been accounted for the whole 10-year period.
Costs are all estimated in today’s value (2021 EUR) and as incremental, relative to the estimated
baseline costs. The baseline costs are the costs likely to be incurred in the absence of a new
regulation, or in the continuation of the current situation, assuming all aspects of the current legal
framework of European population statistics remain unchanged.
A 3% social discount rate has been applied, in line with the recommendations of the Better
Regulation Guidelines.
19.3. 3.3. Labour costs
With regards to the NSIs, estimations of the different labour costs across each Member State were
made using the 2016 labour cost of the public administration sector113
. Annual labour costs have
been estimated up to 2021 in line with the HICP AARoC114
. Between 2021 and 2023 included were
estimated in line with the OECD HICP forecast. Annual labour costs from 2024 onwards were
estimated based on the average of the annual rate of change applicable between 2021 and 2023 for
each Member States and assumed constant over for the remaining of the study period.
In order to obtain daily labour costs projections across each member States, it was assumed that 1
FTE corresponds to 215 working days per year, as follows:
Estimated Daily NSIs staff labour costs = {PA estimated annual labour costs /215}.
Eurostat labour costs were estimated on the assumption that a monthly salary of an AD10 official
was EUR 9,000 and using 215 working days per year (equivalent to 17,9 working days per month),
as follows:
Estimated Daily Eurostat staff labour costs = {AD10 officials monthly salary /17.9}.
19.4. 3.4. Costs estimation
Most of the cost estimates were produced using the following formula:
N. of FTEs * Level of effort (person days) *Daily Labour Costs,
based on data points and costs estimates provided through the consultation process (feedbacks from
Eurostat, interviews and NSIs surveys, Public Consultation).
113
Labour cost, wages and salaries, direct remuneration (excluding apprentices) by NACE Rev. 2 activity) - LCS
surveys 2008, 2012 and 2016 [lc_ncost_r2], available at:
http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=lc_ncost_r2&lang=en
114
HICP - annual data (average index and rate of change) [PRC_HICP_AIND__custom_1686993], available at:
http://appsso.eurostat.ec.europa.eu/nui/show.do?dataset=prc_hicp_aind
128
20. 4. IMPACT ASSESSMENT
20.1. 4.1. Baseline scenario
In the baseline scenario A, limited harmonisation of population definition will remain, statistical
processes will remain separate, and the current requirements in terms of statistical outputs will
continue as they are.
The cost assessment of the baseline scenario was carried out based on the evaluation results. For all
sub-measures included in policy option A, the status quo will remain. Notwithstanding the sunset
clause in Regulation (EU) No 1260/2013, policy option A assumes that a replacement legal act on
demography statistics will be proposed when it expires in 2028.
It is important to note that the costs estimated for policy option A (the baseline scenario) are not the
same as the costs associated with the current legal framework. Indeed, all recurrent costs incurred
by Regulation (EC) No 862/2007 (on migration statistics), No 763/2008 (the Census Regulation)
and Regulation (EU) No 1260/2013 (demography) were estimated to remain constant and
calculated over a 10-year period for both the European Commission (Eurostat) and the 27 Member
States (EU-27) and their respective NSIs.
However, in addition to these costs, costs associated with the provision of voluntary statistics,
maintaining statistical population registers (for those Member States that currently have or are
planning to implement one regardless of this initiative) and maintaining a strict Usual Residence
definition of population (for those Member States that currently adopt this definition) are also
included in the baseline costs. The rationale for the inclusion of these costs in the baseline is that:
 It was assumed that those Member States with a population register115
have incurred and
will continue to incur costs relating to keeping these registers up-to date. These costs were
therefore accounted for in policy option A. Similarly, some NSIs are planning to transition
to a register-based system116
in the near future and will therefore incur costs. As these
Member States are planning to use national statistical registers regardless of a potential new
regulation, the forthcoming costs cannot be attributed to a new regulation that would include
a provision requiring for NSIs to set up and maintain national statistical registers.
 Some Member States are already strictly using the population definition based on the usual
residence concept117
. Those Member States are already incurring costs relating to the use of
this definition and will continue to do so. These costs cannot therefore be attributed to a new
regulation that would include a provision requiring NSIs to either strictly use the usual
residence definition or to restrict their possibilities to use the legal or registered population
definitions.
 Some Member States are already providing data on a voluntary basis. Data currently
collected on a voluntary basis incur costs to both the European Commission (Eurostat) and
115
NSIs survey responses (Q. ‘Does your NSI maintain a national statistical register?’. The following NSIs responded
‘Yes’: Austria, Belgium, Bulgaria, Denmark, Estonia, Finland, Italy, Latvia, Netherlands, Slovakia, Slovenia, Spain,
Sweden).
116
NSIs survey responses (Q. ‘Does your NSI maintain a national statistical register?’. The following NSIs responded
‘No, but planned’: Croatia, Cyprus, France, Hungary, Lithuania, Malta, Portugal).
117
NSI survey responses (Q. ‘Which definition(s) of the population is used by your NSI?’. The following NSIs
responded ‘strict usual residence’: Bulgaria, Croatia, France, Hungary, Ireland, Lithuania, Malta, Portugal, Romania).
129
NSIs. These costs are likely to remain irrespective of whether a new regulation makes the
existing voluntary data mandatory
Therefore, the costs attributable to the baseline scenario and incurred by both the Commission
(Eurostat) and the NSIs can be categorised as follow:
 Costs of mandatory migration statistics
 Costs of mandatory demography statistics
 Costs of mandatory Census (population and housing statistics)
 Costs of maintaining Statistical Registers
 Costs of maintaining harmonised population definitions
 Costs of Voluntary Statistics
The costs and benefits estimated as part of this impact assessment (qualitatively and quantitatively)
are incremental costs/benefits, and therefore presented as additional costs/benefits relative to the
baseline scenario A.
Details and values of the estimated costs for the baseline are included in the main report.
20.2. 4.2. Policy options and factors considered for assessing costs
As mentioned earlier, the policy options have been built and therefore assessed considering three
main factors or criteria: national statistical registers, levels of further harmonisation in population
definitions and the level of statistical outputs.
20.2.1. 4.2.1. National statistical registers
The requirement for NSIs to implement national statistical registers of measure 1.2118
is only
introduced in policy option D.2 (the status quo being maintained under all other options).
Therefore, for all options except D.2, no incremental (additional relative to the baseline) costs will
be incurred due to this factor.
Under policy option D.2, all Member States that do not currently have a register and are not
planning on having one119
, will incur costs resulting from the obligation of setting up a statistical
register.
Details and values of the costs incurred by these Member States and the Commission are included
in the main report.
118
Measure 1.2: NSIs establish national statistical registers.
119
NSIs survey responses (Q. ‘Does your NSI maintain a national statistical register?’. The following NSIs responded
‘No, and not planned’: Germany, Ireland, Luxembourg, Poland, Romania, Slovenia).
130
20.2.2. 4.2.2. Costs savings for the NSIs due to integrated statistical processes
An estimate of the cost savings to Member States in the long run associated with the delivery of
register-based censuses, relative to traditional or combined censuses, was estimated for those
Member States that currently do not and are not planning already to implement a population
register. Costs saving are expected to occur because of integrating statistical processes through the
implementation of statistical population registers. In this analysis, costs saving were estimated
based on data on the total cost of delivering the 2000/01 and 2010/11 census rounds by census
methodology available from the UNECE120
. Costs provided for Member States were converted into
2021 EUR and then the average and median per capita cost of delivering each census round across
all Member States for which data was available were calculated, by census methodology. The
difference between the cost of delivering a census overall (i.e. across all census methodologies),
and the cost of delivering a register-based census was then calculated121
. Based on this, the per
capita cost saving associated with a register-based census was estimated to be EUR 4.52 in 2000
and EUR 5.05 in 2011. Adopting a conservative approach, the lowest figure was selected, and a
10% error range applied to this122
. Maintaining this conservative approach, the lower range figure
for the per capita cost saving was then used to estimate total savings to Member States123
.
To estimate the cost savings per capita over time, national population in the EU as of 1st
January
2021124
, and population projections125
for 2025, 2030, 2035 from Eurostat have been used. To have
yearly projections, a simple linear trend function has been applied to cover for the years missing
between 2021-2025, 2025-2030, and 2030.
The total estimated costs savings for the relevant NSIs are around EUR 575 million.
20.2.3. 4.2.3. Harmonisation of population definition
The options are introducing two different levels of further harmonisation of population definitions
relative to the Baseline (Policy option A)
In policy Option B.1 and B.2, like in the baseline, the harmonisation of population definitions
remains fragmented. There are therefore no incremental costs associated with this factor for policy
option B.1 and policy option B.2.
120
UNECE (2014) Measuring population and housing - Practices of UNECE countries in the 2010 round of censuses;
available at https://unece.org/DAM/stats/publications/2013/Measuring_population_and_housing_2010.pdf and UNECE
(2008) Measuring population and housing - Practices of UNECE countries in the 2000 round of censuses, available at:
https://unece.org/DAM/stats/publications/Publication_on_2000_censuses.pdf.
121
The median was selected as the preferred metric since it is less sensitive to outliers.
122
i.e. to allow a margin of error, the per capita cost saving was estimated to range from between EUR 4.07 to
EUR 4.97 (+/- 10% of 4.52).
123
This was estimated as a figure of EUR 4.07 per capita. In addition, it was noted that three Member Stated appeared
to have changed census methodology between the two rounds, of which two changed to a register-based census. The
cost difference between the delivery of the 2000/01 and 2010/11 census rounds for those Member States was also
calculated, as a possible proxy/ further check on the estimate of per capita cost savings associated with switching to a
register-based census. This cost saving associated with a register-based census, relative to a traditional or combined
census, was estimated to be EUR 6 per capita. Note, however, that this result must be interpreted carefully: the estimate
was calculated based on data from only 2 Member States and does not account for other factors that may have affected
costs over time (i.e. between the two census rounds).
124
https://ec.europa.eu/eurostat/databrowser/view/tps00001/default/table?lang=en
125
https://ec.europa.eu/eurostat/databrowser/view/PROJ_19NP/default/table?lang=en
131
In policy options C.1 and C.2, harmonisation of population definition is improved (measure 2.3)126
.
Due to the limitation in the definitions of what “justified exemptions” entail, it was assumed that all
Member states currently not already using a definition strictly based on the usual residence concept,
and/or not using the legal or registered residence would have to completely switch to using the
usual residence concept for defining their population base127
. This is based on the assumption that
those 8 Member States (using either combined or other definitions) are likely to no longer be
eligible for the justified exemptions.
Policy options D.1 and D.2 introduce a full harmonisation through the measure 2.2128
. The costs of
fully harmonising the population definitions based on the strict usual residence concept are going to
be borne by all 17 Member States129
that are not currently using this definition for all datasets.
Details and values of costs associated with the different levels of harmonisation of population
definitions are set out in the main report.
It should be noted that for this factor, the Commission (Eurostat) and the Member states will incur
similar costs in policy options C.1 and C.2, associated with the harmonisation improvement.
Similarly, costs associated with the full harmonisation will be equal under policy options D.1 and
D.2.
Limitation should be noted around the costs’ estimation and the feasibility for certain Member
states to use the strict usual residence concept to define their population base. For the purpose of
this study, it was assumed that the level of effort and number of days required for such a transition
would be the same across all Member States and that transitioning to a strict definition was possible
for every Member States in terms of technical and operational feasibility. However, a review of the
feasibility studies carried out by the Member States in 2016 has demonstrated that this would not
necessarily be the case for all Member States130
. Nevertheless, methodological work has progressed
in the meantime, for instance some Member States have advanced on more sophisticated register-
based methods around ‘signs of life’ that could ultimately allow efficient adjustments of the
population base at microdata (person record) level.
20.2.4. 4.2.4. Statistical outputs
The options are introducing three different levels of ambition relative to the baseline (policy option
A) regarding statistical outputs requirements.
 Policy option B.1 and policy option C.1 introduce the smallest update in terms of statistical
outputs. Updated statistical requirements will be introduced in demography, migration, and
126
Measure 2.3: harmonised population definition (based on usual residence concept) with stricter and justified
exceptions.
127
NSI survey responses (Q. ‘Which definition(s) of the population is used by your NSI?’. The following NSIs
responded ‘combined or other’: Austria, Czechia, Estonia, Germany, Latvia, Poland, Slovenia, Spain).
128
Measure 2.2: Strictly harmonised population definition (based on usual residence) for all datasets.
129
NSI survey responses (Q. ‘Which definition(s) of the population is used by your NSI?’. The following NSIs
responded ‘legal residence’, ‘registered residence’, ‘combined’ or ‘other’: Austria, Belgium, Czechia, Denmark,
Estonia, Finland, Germany, Italy, Latvia, Luxembourg, Netherlands, Poland, Slovakia, Slovenia, Spain, Sweden).
130
Feasibility studies on the use of strict usual residence concept for population definitions carried out according to
Article 8 of Regulation (EC) No 1260/2013. Denmark mentioned that such a transition would be very costly, while
Germany assessed this measure as not feasible.
132
population and housing census statistics on statistical topics that are covered by the current
EU legislation. No new statistical topics and outputs will be implemented.
 Policy option B.2 and C.2: Statistical topics covered by the current EU legislation will be
further extended relative to requirements in policy options B.1 and C.1. Some existing
voluntary statistical topics will become mandatory. Upgraded statistical requirements will be
introduced in demography, migration, and population and housing census statistics. New
statistical topics will be added
 Policy options D.1 and D.2: Statistical topics covered by the current EU legislation will be
even further extended relative to requirements in policy options B.2 and C.2. Some of the
voluntary statistical topics will become mandatory. In addition, new statistical topics will be
added. Statistical output will reach most ambitious level by covering existing and new topics
of demography, migration, and population and housing census statistics. Some of census
outputs will be required more frequently than every 10 years
Analysis of these three levels of ambitions in terms of statistical outputs requirements has been
carried out to understand which policy option contain measures on statistical outputs that simply
entail an incremental level of effort relative to the baseline (i.e. the current regulation) by requiring
the same type of datasets but in more details and/or characteristics and which policy option, instead
would require the introduction of entirely new datasets and/or making voluntary dataset mandatory.
Depending on each of these levels of ambition, incremental levels of effort were estimated
proportionately and relative to the baseline by estimating the additional days and FTE required for
meeting the different requirements brought by the various measures and their level of ambitions.
Incremental costs were estimated based on the same assumptions and data drawn upon to estimate
costs in the evaluation. These costs were associated with administrative costs to Member States,
and enforcement costs to Eurostat, since it is assumed, these are the only cost items that will
increase relative to the baseline.
Details and values of estimated costs incurred by both the Commission (Eurostat) and the Member
states and resulting from these different levels of ambition in terms of statistical outputs are set out
in part 1.1 of Annex 2. It should also be noticed that these costs borne by the Commission
(Eurostat) and the Member States due to the least ambitious upgrade are therefore equal in policy
options B.1 and C.1; Cost incurred by a more extensive upgrade of statistical outputs are similar in
policy options B.2 and C.2; and costs incurred due to a major and most ambitious upgrade of
statistical outputs are equal in policy options D.1 and D.2.