Overview of Approaches to Register-Based Populating Censuses

1
Overview of Approaches to Register-Based
Populating Censuses
Regional Workshop on the 2010 World Programme on
Population and Housing Censuses Minsk, 8-12
December 2008

• Paolo Valente
• United Nations Economic Commission for Europe
• Statistical Division

2
Content

1. General considerations about using registers
2. Fully register based census
3. Data from registers full enumeration
4. Data from registers existing sample surveys
5. Data from registers ad-hoc sample surveys
6. Conclusions

3
1. General considerations

• Advantages of using register data for censuses
• Use information already available ? No additional
  burden on respondents
• Data potentially available every year
• Cheaper that full enumeration once high quality
  registers are established

4
1. General considerations

• Requirements of using register data
• Need to have public and legislative support
• Long time and large investments needed to develop
  high quality statistical register system to be
  used for censuses

5
2. Fully register based censuses

• Census conducted by integrating data from various
  registers
• Conclusion of long step-by-step process
• Census is part of a system of register-based
  information

• Main requirements
• Availability of population register and dwelling
  register
• Capacity to link persons and dwellings,
  identifying households
• Public and legislative support for the use of
  population registers

6
2. Fully register based censuses

• Experience 2000 census Finland, Denmark,
  Iceland, Norway
• From 2010 Sweden, Netherlands, Slovenia
• Advantages
• No burden on individuals
• Data usually available every year
• Once high quality registers are established, it
  is cheaper than a full field enumeration
• Limitations
• Setting up and maintenance require significant
  resources
• Characteristics to be collected are limited to
  those on the registers
• Quality of census depends on coverage and quality
  of registers
• Data in registers usually collected for
  non-statistical purposes
• Statistical agencies are also dependent on
  register authorities

7
3. Data from registers full enumeration

• Option when registers exist but some important
  variables are missing, or quality is not
  sufficiently good
• Enumeration provide data on these variables, and
  evaluation of coverage and accuracy of register
  data
• Some register data can be pre-printed in
  questionnaires

8
3. Data from registers full enumeration

• 2000 round Belgium, Latvia, Slovenia, Spain,
  Switzerland
• Advantages
• Data collection operations could be used to
  update and improve registers
• Preliminary step to  pure  register-based
  census
• Distinction between statistical operation
  (census) and admin. operation (register update)
  should be clear to the respondents
• More efficient than traditional census in field
  operations
• Disadvantages
• Method can be complex to apply
• More expensive than pure register based census
• Burden on the public

9
4. Data from registers existing sample surveys

• As an alternative, register data can be
  integrated with data from existing sample surveys
  (i.e. LFS)
• Survey results are used for variablesnot covered
  in registers
• Data linked at individual level

10
4. Data from registers existing sample surveys

• 2000 round Netherlands ( Virtual census )
• Advantages
• No burden on respondents
• Reduced costs no specific field data collection
• Consistency between census results and survey
  results
• Disadvantages
• Process to successfully link information on
  individuals from different surveys and registers
  is complex
• Limited detail for information taken from sample
  surveys

11
5. Data from registers ad-hoc sample surveys

• Other alternative using ad hoc sample surveys
  instead of existing surveys.
• Ad-hoc surveys can be used to evaluate accuracy
  of registers and collect traditional long form
  census data.
• Requires capacity to link individual data from
  different sources

12
5. Data from registers ad-hoc sample surveys

• Plans for 2010 Austria, Belgium, Germany,
  Israel (2008), Switzerland
• Advantages
• Accuracy of population register can be tested
• Population counts could be adjusted
• No burden on respondents
• Reduced costs no specific field data collection
• Disadvantages
• Method could be complex
• Limited detail for information taken from sample
  surveys

13
Conclusions Census methods adopted in ECE
Region2000 census round and plans for 2010 round
countries
countries
14
Conclusions

• UNECE countries seem to find their way to use
  data they have in registers for the census
• Quality of register data is a key factor, but not
  an  excuse  not to use them
• Learn from other countries experiences but then
  take decisions based on national context

15
Source on Register based censuses

• More details available in CES Census
  Recommendations, Appendix II

http//www.unece.org/stats/census/
16
Sources on Register based censuses

• Fully register-based censuses also covered in
  UNECE publication on register-based statistics

http//www.unece.org/stats/census/
17
Other UNECE material on censuses

• CES Recommendations for 2010 Censuses
  (English/Russian/French)
• Documents of annual meetings on censuses
• Documents of training workshops on censuses
• Methodological and reference material
• Collection of questionnaires and material used in
  2000
• Publication on national practices in 2000
  censuses (Eng/Rus)
• 5 multimedia presentations on censuses (Eng/Rus)
• http//www.unece.org/stats/census/