4th International Conference on Population Geographies The Chinese University of Hong Kong 1013 July

1
4th International Conference on Population
GeographiesThe Chinese University of Hong Kong
10-13 July 2007Session 4A International
MigrationThe settlement patterns of the
foreign population in Italy at the start of the
21st century.Corrado Bonifazi, Frank Heins
and Salvatore Strozza Institute of Research
on Population and Social Policies, National
Research Council, Rome, Italy University of
Naples Federico II, Naples, Italy
2
Structure of the presentation

• Aim
• The foreign population in Italy
• The dissimilarity index
• Regional patterns of dissimilarity
• The possible determinants

3
Aim of the project

• The aim of our project is the (1) analysis of
  the regional settlement patterns of the foreign
  population in Italy in a national perspective and
  (2) the comparative analysis of its local
  settlement patterns.

4
Aim of the presentation

• The following presentation focuses on the
  comparative analysis of the local settlement
  patterns. The local labour market areas (686
  areas defined based on the commuting patterns
  found during the 2001 population census) are the
  reference areas and the observations are the
  single census tracks.
• Results regarding the dissimilarity index,
  measuring the dimension of evenness in the
  distribution of the foreign population compared
  to the Italian population, are presented.

5
The foreign population in Italy

• The foreign population amounted to 1334889 at
  the 2001 population census. Since then it grew
  rapidly.

6
The foreign population in Italy

• the share of the foreign population at the
  census in 2001 was 2.3 and stands today at 4.8
  of the total population of Italy
• the foreign population is concentrated in
  Central and North-Eastern Italy the Third Italy
  

7
The dissimilarity index

• Problem the number of foreign residents
  influences the value of the index the lower the
  number of foreign residents the more it is likely
  that their settlement pattern is different from
  the pattern of the Italian population. An aspect
  of aleatory to the value of the index is
  introduced.
• We refer to the review of dissimilarity indices
  by Massey and Denton and the more recent
  contributions by Apparicio et al and Wong, who
  introduced GIS to the calculation of the
  dissimilarity index.

8
The dissimilarity index

• The dissimilarity index is defined in the
  following way
• with
• i as the indice of the territorial unit (in our
  case the census tracks)
• psi foreign population of the census track i
• PS foreign population of the Local Labour
  Market area of reference
• pii Italian population of the census track i
• PI Italian population of the Local Labour
  Market area of reference

9
The modified dissimilarity index

• To standardise the territory we applied the
  concept of composite population counts (see
  Wong) the observed population date are
  substituted through a weighted mean of the
  population data of the adjacent census tracks.
  The distance between the centres of the census
  tracks (in km) serves in the present case as a
  straightforward weighting model

10
The modified dissimilarity index

• This simple distance criteria does not take into
  consideration natural or man-made barriers, which
  would be important to determine social
  interactions between adjacent areas. The distance
  parameter expresses a hypothesis regarding the
  effect of distance on social interactions the
  higher the exponent a, the smaller is the area
  taken into consideration to calculate the value
  of the composite population counts. To test the
  effects a was set to 1, 2, 4, 8, 16, 32, 64 and
  128. This approach could be a step towards
  tackling the modifiable areal unit problem.

11
The modified dissimilarity index
12
The modified dissimilarity index

• Before proceeding with the calculation of the
  modified dissimilarity index the contributions to
  the composite population counts of census track i
  of each census track j were re-proportioned to
  its observed population numbers. This correction
  is necessary to make sure that the population of
  a census track is not taken into consideration
  more often than others only due to the small
  distance between census tracks especially in the
  case of city centres.

13
The dissimilarity index
14
The not-modified dissimilarity index

15
The modified dissimilarity index

• CPC and distance parameter a 8

16
The modified dissimilarity index

• CPC and distance parameter a 2

17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
Conclusion

• a low proportion of foreign residents, high
  unemployment and a low gdp per capita are
  associated with an uneven distribution of the
  foreign population
• it does not seem obvious to use the
  dissimilarity index, even in its modified form,
  in a comparative study of the local settlement
  patterns of the foreign population in Italy