The Development and Use of Registers in Finland

1
The Development and Use of Registers in Finland

• Royal Statistical Society, Statistics User Forum
  Annual Conference
• 16 November 2006, Riitta Harala, Director,
  Population Statistics

2
Statistics Finland

• National Statistical Office
• Founded in 1865
• Responsible of 77 per cent of the production of
  official statistics
• Under the Ministry of Finance
• Personnel about 1000 persons
• 6 statistics production departments, 3 supporting
  departments

3
Register-based data collection

• Long experience, since the 1970 population census
• first in social and demographic statistics,
  nowadays more and more also in business
  statistics
• about 96 of input data comes from
  administrative sources (as measured in number of
  statistical units times number of variables)
• direct data collection also very inportant
• these two methods are complementary to each other

4
Register-based statistics production Key
preconditions

• Legal basis
• Acceptance of the population
• cost efficient
• reduces response burden
• Unified identification code systems
• Comprehensive and reliable register systems
  developed for administrative needs
• Cooperation among administrative authorities

5
Ways of utilising administrative data

• Direct use of register data
• Register estimation
• Combined use of survey and register data
• additional information from registers
• use as sampling frame
• non-response control
• imputation
• determining the structure of non-response

6
Examples of totally register-based statistics in
Statistics Finland

• Population statistics
• Population censuses
• Building and dwelling statistics
• Statistics on housing conditions
• Regional employment statistics
• Statistics about justice and crime
• Election statistics
• Income statistics

7
Examples of surveys in which register data are
used

• Income distribution statistics
• Household budget survey
• Labour force statistics
• Wage and salary statistics
• Enterprise statistics

8
Authorities delivering administrative data

• Population register centre
• Taxation authorities
• Central Pension Security Institute
• The State Treasury
• Municipal Pension Institution
• National Pension Institute
• Ministry of Labour
• Ministry of Justice etc.

9

• The best example of the use of administrative
  data sources in Statistics Finland is the
  register-based population and housing census
  system

10
Development of the use of registers in Population
Censuses

• took 20 years
• step by step
• full support of the Ministry of Finance and the
  management of Statistics Finland

11
Use of administrative register data in Population
and Housing Censuses 1950-2000
12
Most important reasons for beginning to build a
totally register-based census system

• costs
• administrative registers were sufficiently
  developed
• growing need for statistics

13
(No Transcript)
14
Estimated costs/person of the censuses in 2000
(in euro)
15
Statistical basic registers
16
Units belonging to the Register-based Population
Census System
and links between units
Map coordinates
Register of Buildings
Buildings
and Dwellings
Business
(Population Register
Summer
Dwellings
Centre)
cottages
premises
House-
Population Register
hold
(Population Register
dwelling
Centre)
units
Families
Persons
Business Register
Enter-
Establish-
(Statistics Finland)
prises
ments
17
Examples of the registers used in defining the
main type of activity of persons

• Population information system
• Register of Job Applicants
• Pension registers
• Study Aid Register
• Joint Selection Register
• Private sector employment register
• Central government sector employment register
• Local government sector employment register
• Social Insurance Institutions employment
  register

18
Cont.

• Evangelical Lutheran Church employment register
• Bank of Finland employment register
• Provincial Government of Åland employment
  register
• University of Helsinki employment register
• Employment registers of certain major cities
• Register of Conscientious Objectors
• Register of Conscripts
• Personal Tax Register
• Customer database

19
Cont.

• Employers annual notifications
• Wage statistics
• Register of Completed Education and Degrees
• Student Register
• Business Register

20
Advantages of register-based statistics production

• Total coverage for lower costs (1980 Finnish
  Census cost EUR 34 million, 2000 Census EUR
  0,84).
• Because of the identifying codes it is possible
  to compile the flow-statistics
• Possibility to produce more data (for instance
  multiple activities)
• Possibility to link with other material
• No subjective opinions gt All the cases are
  handled in the same way
• No additional burden to respondents
• Fewer personnel problems
• Confidentiality advantages

21
Disadvantages of register-based statistics
production

• Only variables covered by registers are available
  
• New variables by combining data from several
  registers
• Some restrictions on definitions of units and
  variables
• Example Dwelling households only
• Vulnerable to changes in legislation and
  practices
• Census no longer a tool for ad hoc data
  collection
• Dependence on data suppliers

22
New opportunities and challenges

• Release of data for research purposes
• New kinds of statistics
• Small area statistics and GIS

23
Release of micro data for research purposes

• Register-based databases are the goldmines of
  empirical research, for instance sociological,
  health and economic analysis
• The challenge protection of confidentiality vs.
  providing researchers with the relevant data
  cost-effectively
• according to the Statistics Act confidential data
  may be released for purposes of scientific
  research in unidentifiable form

24
Forms of release of micro data

• Off-site release of unidentifiable samples to
  outside researchers, about 200 cases a year
  (mostly population statistics
• On-site Statistics Finlands research
  laboratory, mostly business statistics data and
  combined employee-enterprise data (about 20 cases
  a year)
• Plans for on-line researcher services

25
Data files for research purposes

• Longitudinal data files (from which samples can
  be released to researchers)
• Longitudinal data file on population and housing
  censuses 1970-2005
• longitudinal data file on employment statistics
  1987-2003
• longitudinal data file of family formation
  1970-2003
• longitudinal data file of the impacts of labour
  policy measures 1987-2003
• longitudinal data file of living conditions and
  causes of death (1970-2003)

26

• Only to be used in the Research laboratory
• FLEED data file Finnish Longitudinal
  Employer-Employee Data
• Financial statements panel
• R D panel
• ICT panel
• Industrial statistics panel

27
New kinds of statistics

• Flow statistics
• linking unit data from consecutive years
• possibility to describe how peoples activities
  and conditions change over time, for instance
• the entry of graduates into the labour market
  (placement statistics)
• the development of employment of immigrants
• the rise of social status
• changes of jobs or labour market status etc.

28
Immediate transition to education of
matriculation examination graduates 2004
29
(No Transcript)
30
Employment of graduates year after graduation by
level of education 2003 and 2004
31
(No Transcript)
32
(No Transcript)
33
Small area statistics and GIS

• Traditionally the most important regional unit in
  statistics has been the administrative area
• map coordinates for buildings were collected in
  the 1970 census and has been maintained since in
  the Population Register Centre
• the Finnish register-based statistical system is
  point-based rather than area-based

34

• the link connections
• person -gt dwelling -gt building -gtmap coordinate
• employed -gt enterprise -gt establishment -gt
  building -gt map coordinate
• excellent basis for small area statistics and use
  of GIS

35
Buildings coordinates are aggregated to
grids. This enables all data linked to buildings
to be aggregated to grids.
36
The threat of depopulation in Pirkanmaa by sq.
km grids
Sq. Km grid where the youngest inhabitant is
aged 50 or over
Sq. Km grid where the youngest inhabitant is aged
under 50
37
Industrial work places in Lohja by sq. km grids
38
The population in Lohja by 250 m x 250 m grids
39
The population in the impact area of of a
motorway by 250 x 250 meters grids
40

• Use of Registers and Administrative Data Sources
  for Statistical PurposesBest Practices of
  Statistics Finland
• Can be downloaded in pdf-format in
  addresshttp//tilastokeskus.fi/censusbyregisters