Click here for main title Click here for subtitle

1
Migration, Areas, and Social Change
Ian Shuttleworth School of Geography,
Archaeology and Paleoecology Queens University,
Belfast
Northern Ireland Longitudinal Study Information
Morning December 8th 2006
2
Background

• It is well known that migration is a socially
  selective process
• The Laws of Migration Ravenstein (1883)
• Migration is also selective by economic status
  eg Kitching (1990) at an inter-regional scale
• Focus on internal migratory moves in NI post 2001
  by economic status
• Questions
• What kinds of people migrate within Northern
  Ireland
• Does the social selectivity of migration widen
  social divisions between different sorts of areas?

3
Background

• Implications for social policy and spatial
  targeting of resources
• Policy working uphill against social
  selectivity of migration?
• Hitting a moving target?
• Benefiting areas or people?
• Prior expectations, since migration is known to
  be selective in a variety of applications, is
  that internal migration in NI will be socially
  selective and will sift the population
• Is this the case?

4
Data and Method

• Data from the NI LS were used for a preliminary
  examination of these issues
• Information on internal migratory moves within NI
  from 2001 onwards between 890 NI SOAs
• Complex data possibility of multiple moves 890
  890 cell matrix
• So simple descriptive analysis to reduce
  complexity
• Ranks sending and destination SOAs by Noble
  General Deprivation score 10 classes
• Deals only with first moves majority of movers
  have moved only once

5
Data and Method

• The analysis describes
• movements within and between groups of SOAs by
  economic status in 2001
• Considers all movers (eg those within SOAs) and
  movers between SOAs
• Looks at the distribution of movers (by economic
  status) across the 10 classes of SOAs prior to
  and after moving
• Information on individuals (economic status and
  mobility) linked to information on the SOA
  (General Deprivation Score) in which they live
  before and after moving)

6
Data and Method

• Limitations were
• Not all data had been incorporated in the LS
  (only 4-6 tranches) at the time of the
  preliminary analysis
• Individual and household information available
  solely from the 2001 Census no further
  information on changes of status post-2001 (and
  changes of economic status (eg gaining work)
  could be related to internal moves
• No information about causality
• Inferential limits but some tentative
  conclusions can be drawn

7
Results

• Around 20 within the NI LS were classed as
  movers post 2001 across the 10 classes of SOA
  (1most socially deprived.10least deprived) no
  evidence of mobility gradient by social
  deprivation of SOA of original residence
• Most people migrating into the most deprived SOAs
  (Class 1) come from similar types of area
• Most migrating from the most deprived SOAs (Class
  1) go to similar kinds of area
• Highest frequency of moves along diagonal
  between SOAs of the same type

8
Origins of Employed Movers to Class 1 SOAs
9
Flows from Class 1 Deciles (moves between SOAs)
10
Results

• Simple index ratio of observed over expected
  flows between cells (expected column total
  row total divided by grand total)
• Numbers greater than 1 indicate observed flows
  greater than expected less than 1 less than
  expected
• Migration Stickiness movers stay more than
  expected in SOAs of the same type if anything
  this applies marginally more to those who were
  employed in 2001

11
Migration Stickiness Those Employed in 2001
(external flows)
12
Migration Stickiness Those Inactive in 2001
(external flows)
13
The Social Composition of Deciles prior to and
after moving

• The most deprived classes of SOA (classes 1
  through 4) experienced net out-migration Classes
  5 through 10 saw net in-migration
• This was generally observable across all economic
  statuses
• There is some evidence that the movers
  redistribute themselves differentially across the
  10 SOA classes so that the proportion of employed
  movers in deprived areas declines and that of the
  inactive grows
• This is likely, however, to have only a small
  effect on the overall social composition of areas
  (since only c. 20 are classed as movers)

14
(No Transcript)
15
Results

• Clear differentials between the SOA classes in
  economic status of movers both before and after
  moving more inactive in Class 1 fewer employed
  in Class 1
• Internal migration makes some minor changes but
  the big picture remains in the differentials
  overall
• Need to control for distance SOAs in same type
  spatially clustered?

16
Conclusion

• Limitations in the analysis
• No data on changes of economic status at time of
  movement possible (probable?) that
  losing/gaining a job is correlated with moving
• Economic status as measured in 2001 could not be
  expected to be a good measure of status in 2006
  so analysis tied to years around the Census?
• Future opportunities
• Possibility of adding data on benefit databases
  entering or exiting the claiming of benefits and
  relationship with mobility
• Using more tranches of data

17
Conclusion

• Future opportunities
• Rich data (more information on other household
  members other types of migrants other
  variables)
• Can augment and add extra dimensions to analyses
  beyond the standard output from the Census (eg
  flow data) Fieldings analyses of
  inter-regional migration in England Wales using
  LS