Migration

1
Migration commuting substitution Commuting
potential and interregional migration propensity
Emma LundholmUmeå UniversitySwedenEmma.Lundholm
_at_geography.umu.se
2
Development in Sweden and other European
countries.

• Migration tolerance has decreased
• Commuting tolerance has increased

3
Is increased commuting a good thing or a bad
thing?
4
Interregional migration in Sweden
5
Substitution hypothesis

• people are today more likely to choose commuting
  over other strategies, including interregional
  migration

6
Substitute or complement?(Evers Van der Veen,
1995)

• It is more convinient to commute
• Migration is redundant
• Enables a more free choice of settlement
• It is more difficult to migrate
• Commuting is the solution to achieve labour
  market matching

7
Previous studies have shown..

• People who live in regions with good commuting
  potential are less likely to migrate

Eliasson, Lindgren, and Westerlund 2003 Eriksson,
Lindgren, and Malmberg 2007 Van Ham, Mulder, and
Hooimeijer 2001
This is a study of this relationship over time
8
Empirical question

• Have migration propensities declined more in
  regions with better commuting potential?
• cross-sections 1970, 1985, 2001

9
Commuting potential size of labour market
Approximated as population at a given eucledian
distance
10
Method and Data

• Register data, entire Swedish population in
  working age (1970, 1985, 2001)
• Interregional migration migrants moving gt150
  kilometer
• Logistic regression

Commuting potential narrow labour market
extensive labour market
Dependant variableInterregional migration
propensity
Control variablesage,sex,civil
status,children in household,education
level,employment,student, recent migration
11
Results

• Migration was less likely among persons living in
  regions with better commuting potential
• The inhibiting effect of residing in a large
  labour market was the same in all three years
• no support for the hypothesis that commuting
  potential reduce interregional migration more
  today

12
1970 1985 2001
Woman -0,059 0,114 0,086
Age 18-20 -0,167 -0,235 -0,145
Age 21-24 0,184 0,166 0,226
Age 30-34 -0,314 -0,325 -0,390
Age 35-44 -0,767 -0,653 -0,815
Age 45-54 -1,358 -1,234 -1,545
Age 55-65 -1,936 -1,644 -1,912
Married -0,109 -0,194 -0,174
Children -1,245 -2,030 -2,435
High education 1,127 0,892 0,467
Employed -0,244 -0,226 -0,237
Student -0,231 0,705 0,604
Recent migration 1,017 1,222 0,837
LM size 0-30 km -0,144 -0,158 -0,160
LM size 30-80 km -0,288 -0,265 -0,243
LM size 0-30 km children 0,048 0,142 0,255
LM size 30-80 kmchildren 0,031 -0,003 -0,114
N (included in analysis) 4183145 4585523 5300630
Model chi-square 60001 51448 128219
-2 Log likelihood 658551 499842 774110
Nagelkerke R square 0,090 0,098 0,153
13

• Has increased job commuting substituted
  interregional migration
• or
• has decreased interregional migration forced the
  process of job commuting?

14
Summary

• Commuting potential is a factor that can
  contribute to non-migration
• No direct casual relationship between increase in
  commuting and lower migration rates over time
• Increased migration to commuting substitution
  might not be a general trend but rather a
  tendency among increasingly immobile groups, such
  as dual income households

15
Thank You!