Internal Migration Around the World: Some Empirical Comparisons

1
Internal Migration Around the World Some
Empirical Comparisons
The Dynamics of Populations Large and Small What
Will Tomorrows World be Like? Symposium in
Honour of Professor Phil Rees
Weetwood Hall, Leeds, UK, July 1, 2009

• Martin Bell and Salut Muhidin

Queensland Centre for Population Research School
of Geography, Planning and Environmental
Management The University of Queensland
2
Context
Source Population Reference Bureau,
http//www.prb.org/pdf08/08WPDS_Eng.pdf
3
Background

• Overall significance of migration
• As a demographic event 1 1 1.8 2.5 12
• As a component of population change
• As the fundamental problematic in projections
• As a political and regulatory issue

4
(No Transcript)
5
Background

• Overall significance of migration
• As a demographic event 1 1 1.8 3.5 12
• As a component of population change
• As a fundamental dilemma in projections
• As a political and regulatory issue
• Why view migration in a comparative framework?
• Aids understanding
• Promotes analytical rigor
• Enhances migration theory
• Assists policy development

6
Prior Work

• Five identifiable strands
• General collections Nam (1990), Rees
  Kupiszewski (1999)
• Specific forms of movement Champion (1989), UN
  (2000)
• Particular aspects of mobility Rogers Castro
  (1981)
• Regional studies CEPAL (2007)
• Partial league tables -

7
League Tables of Internal Migration
Long 1991
World Bank 2008
8
Prior Work

• Five identifiable strands
• General collections Nam (1990), Rees
  Kupiszewski (1999)
• Specific forms of movement Champion (1989), UN
  (2000)
• Particular aspects of mobility Rogers Castro
  (1981)
• Regional studies CEPAL (2007)
• League tables - Long (1991) World Bank (2009)
• Limitations
• Ad hoc - no comparative framework
• Varied definitions of migration
• No commonly agreed indicators
• Limited number of countries

9
The British-Australian Study

• 1997-98 project funded by ESRC/ARC UK end led
  by Phil Rees JS, PB, ODW, Marcus B, GJH
• Major outcomes.
• Four dimensions of mobility
• Migration intensity
• Migration distance
• Migration connectivity
• Migration impact
• Battery of 15 indicators 6 3 4 - 2
• Defined computational methods analytical rigor
• Systematic review of impediments to comparisons
• Limited test using data for Britain and
  Australia 6.8 moves cf 12.5

10
Objectives

• Specific Aims
• More comprehensive assessment of OZ-UK study
  measures
• Extend analysis across larger sample of countries
• Explore trends over time
• Triggers
• Request from UN Development Program for
  comparative estimates
• Data availability
• UN HDR Research Paper release today
  (http//hdr.undp.org/en/reports/global/hdr2009/pap
  ers/).
• Focus on
• migration intensity
• age selectivity
• spatial redistribution

11
Data Sources

• Minnesota Population Centre
• IPUMS (Integrated Public Use Microdata Series)
• Census data for 35 countries - 25 with internal
  migration
• National Statistical Offices and UN contacts
• Data for a total of 28 countries O-D matrices
  various geographies

12
Impediments to Comparisons

• How migration is measured
• Type of data (event or transition)
• Coverage and quality of the data
• Interval over which migration is recorded
• 1 year, 5 years, 10 years, lifetime, variable
• Spatial framework used to record a move
• zonal system (MAUP scale and zonation)

13
Census Migration Intervals
Source Adapted from Bell (2005)
14
Migration Interval
Note Data only available from the 1990 census
15
Variation in Zonal Systems
16
Which Measure of Intensity?

• Oz-UK Study - Migration expectancy number of
  lifetime moves
• Requires data for a single year interval
• Long 1991 Crude migration intensity - of
  population who changed residence
• Can compute for any interval, but normally
  reported for all moves
• Courgeau 1973 Migrations et decoupages du
  territoire
• If propensity to move f(distance) then MI f(n
  of zones)
• The finer the spatial mesh, the larger the number
  of migrations that will be captured and hence the
  greater the recorded intensity
• If distance decay is a simple power function,
  then CMIk log (n2)
• Plot CMI against log (n2) .19 countries51
  observations

17
Five Year Migration Scatterplot 19 countries, 51
observations
18
Crude Migration Intensity by Zonal System
Selected Countries
19
Courgeau Isolines Lifetime Migration
20
Courgeau Isolines Five Year Migration
21
Courgeaus k
22
Migration and Human Development
23
Trends in Courgeaus k, 5 Year Migration
24
Trends in Courgeaus k, Lifetime Migration
25
Conclusions (1)

• Marked cross-national variations in migration
  intensity but no consistent regional pattern
• Developed countries high in New World moderate
  in Spain/Portugal
• LAC moderate to high in Andes and Central
  America, lower on east coast
• Asia systematically low except Malaysia
• Africa small sample high S Africa, low East
  and West Africa
• Temporal trends
• Falling five year intensities except China,
  Chile, Portugal
• Lifetime intensities rising at a declining rate
• Loose relationship with HDI

26
Conclusions (2)

• No magic indicator akin to TFR, E0
• Courgeaus k enticing possibilities
• Some basis for systematic comparison
• Allows interpolation
• Harmonises for scale dimension of MAUP
• Facilitates temporal comparisons
• Limitations
• k has no intrinsic meaning
• Highly sensitive to number of observations
• How robust?
• Linear?
• How does k scale? does 2k imply twice the
  intensity?
• Possible elaborations
• Refine distance decay parameters SIMs?
• Simulations based on alternative levels and
  arrangements of national zonal systems?

27
Acknowledgments

• The IPUMS Internationals team at the University
  of Minnesota https//international.ipums.org/inter
  national
• The United Nations Development Program
  http//hdr.undp.org/en/reports/global/hdr2009/pape
  rs/
• Regional statistical offices
• Australia Australian Bureau of Statistics
• China National Bureau of Statistics
• India Registrar General Census Commissioner
• Indonesia BPS Statistics
• Co-author Dr Salut Muhidin
• The University of Leeds
• Phil Rees