Measuring emigration at the census

1
Measuring emigration at the census
United Nations Economic Commission for
EuropeStatistical Division

• Regional Training Workshop on Population and
  Housing Censuses for South Eastern European
  Countries
• Ohrid, 24-26 November 2008

2
Summary of previous episodes

• UNECE analysis of experience of 4 countries at
  2000 Census Round
• Georgia
• Tunisia
• Moldova
• Poland
• Data compared with those from destination
  countries

3
Two critical issues

• Resident population count vs. emigrants count
• Accuracy of emigrants count

4
1. Resident population count

• Accurate count of resident population first goal
  of the census
• Strict application of 12-months rule
• Exclusion of nationals residing abroad from
  population count

5
1. Practices from 4 countries

• Georgia and Tunisia
• separate module
• emigrants excluded from population count
• Moldova and Poland
• questions in main census form
• emigrants included in population count

6
In view of next census..

• Different approaches can be used
• Questions in the main form
• Separate module
• Separate list
• But distinguish
• temporary absent ? IN
• long-term absent ? OUT

7
2. Accuracy of emigration count

• Information is collected on absentees
• Two problems
• Reliability of answers from proxy respondent
• Proxy respondent does not exist

8
2. Accuracy of emigration count

• Three cases
• Entire household (HH) leaves the country
• Some members of HH remain but, before the census,
  the HH dissolves
• Some members of HH remain and fill in the form

9
2. Accuracy of emigration count

• There are cases where proxy respondent does not
  exist ? undercounting
• However, there can also be overcounting

10
Citizens of Poland living in other countries
(2002)
11
Poland emigrants by year of migration
12
Accuracy of emigration count

• Data collection worked well for emigrants that
• Had left the country in the years just before the
  census (up to 5 yrs.)
• Are more likely to keep close ties with the
  country (Polish data were better for Italy and
  Germany than for emigrants in the US or Canada)
• Are members of the family nucleus that is left
  behind (e.g. Tunisian males)
• Are still included in an administrative register
  (Poland)

13
To conclude

• Its fundamental to keep separate the count of
  population from the count of emigrants (persons
  residing abroad for 1 year or more)
• The census cannot provide a good estimate of the
  total number of emigrants living abroad
• Its important to identify what group of
  emigrants we can reasonably count in a census.
  (For example, those having left the country in
  the last years and having close family ties)

14
To conclude

• An emigration module can provide important
  information, not on the total count, but on other
  aspects (geographical distribution of emigrants,
  information on households left behind, reason of
  migration, socio-economic characteristics)
• Accurate test of questions/module is necessary
  (use experience of other countries)
• Choice of respondents is fundamental
  (undercounting vs. overcounting)

15
Emigration topics in the census

• A practice to recommend?
• No accurate count of nationals living abroad
• However
• Strategy to have more accurate population count
• Some useful information on emigration related
  topics

16
Appendix

• Measuring emigration through immigration data of
  receiving countries
• Documentation at
• http//www.unece.org/stats/documents/2008.03.migra
  tion.htm

17
First step

• Identify information needs
• For example
• Total outflows for population estimates
• Outflows of nationals
• Nationals residing abroad (expatriates)
• Short-term emigration

18
Second step

• Map information needs and data sources of
  receiving country
• For example population census can give data on
  stocks of nationals living abroad, not on annual
  flows

19
Third step

• To better use data of receiving country, check
• coverage of the source
• definition of international migrant/migration
• data accuracy
• availability of origin-destination data
• different time references

20
Fourth and final step

• Compare your data with data from receiving
  countries, on flows and stocks
• Do demographic analysis
• Identify weaknesses and strengths of your data
• Possible methods to adjust your data?