F. Domergue - ELIPA : a tool for measuring integration
1. ELIPA : a tool for
measuring integration
Department of Statistics, Studies and Documentation
General Secretariat for Immigration and Integration
Florent Domergue
Roma – 17th
June 2013
2. • ELIPA
• Presentation of the survey
• Characteristics of the new migrants
• Analysis
• Labour market inclusion of the new migrants
• Language training impact for the new migrants
– Summary –
4. – Enquête Longitudinale sur l’Intégration
des Primo Arrivants –
• French public statistical system survey
• Conducted alongside OECD, INSEE, INED…
• Funded by both the Ministry and the European
Commission
• Total cost : 3 M - 3,5 M € (internal costs not included)
5. – Field –
• Migrants who got their first residence permit in 2009,
allowing them to stay in France at least one year (but
likely to stay indefinitely).
• Aged 18 and over
• Coming from « third countries », i.e. outside EEA and
Switzerland
• Signatories of the « reception and integration
contract » (R&I)
6. – Agenda –
• 1st
wave : spring 2010
• 6,107 people interviewed
• 2nd
wave : spring 2011
• 4,756 people interviewed
• 3rd
wave : spring 2013
• 3,540 people interviewed
7. – Aims of the survey –
• Knowledge of the integration path of the new
migrants
• Economy (activity, income)
• « Human capital » (language skills, education)
• Living conditions (housing, family and social life, access to
rights…)
• Evaluation of public policies of reception
• Migratory and administrative processes
• Assessment of the R&I
9. At the time of getting their residence permit :
• 48 % have been in France for less than 2 years
• 16 % for 2 to 4 years
• 36 % for 5 years and over
– Migrants did not settle in France at the
same time –
10. – Main profile : family migration, coming
from Africa –
Economic
migration
Family
migration
Asylum Others Total
All new migrants 9 72 13 6 100
Proportion of women 20 59 42 44 52
Median age 36 32 31 33 32
Average length of stay in France 8 3 3 10 4
Nationality
Europe 5 4 10 3 5
Africa 76 73 38 69 68
Algeria 7 22 1 6 17
Morocco 8 14 0 9 12
Tunisia 5 10 0 3 8
Other african countries 56 27 37 52 32
Asia 12 17 49 23 21
Turkey 2 7 8 6 7
Other asian countries 10 10 41 17 14
America and Oceania 7 6 3 5 6
12. - Migration leads to a breakdown of
employment -
Labour force participation rate
• Before migration : 57 % 2010 : 69 % 2011 : 71 %
Before
migration
2010 2011
Gender
Men 6 24 18
Women 10 48 34
Length of stay in France in 2010
Less than 2 years 8 47 30
2 to 4 years 6 36 28
5 years and over 8 19 17
All new migrants 8 33 24
Unemployment rate (%)
13. - Length of stay in France determines
women’s labour market inclusion -
0
10
20
30
40
50
60
Less than 2
years
2 to less than
4 years
4 to less than
6 years
6 to less than
8 years
8 years and
over
2 years in the labour market
2 years in employment
0
10
20
30
40
50
60
Less than 2
years
2 to less than
4 years
4 to less than
6 years
6 to less than
8 years
8 years and
over
At least 1 year at home
2 years at home
% %
Length of stay in France in 2010 Length of stay in France in 2010
Women in the labour market in 2010 and
2011
Women at least one year at home in 2010
and 2011
14. - For the men, length of stay only impacts
being employed in the long run -
0
10
20
30
40
50
60
70
80
90
100
Less than 2
years
2 to less than
4 years
4 to less than
6 years
6 to less than
8 years
8 years and
over
2 years in the labour market
2 years in employment
Length of stay in France in 2010
Men in the labour market in 2010 and 2011
%
15. - Different timelines with regard to family
and work -
Men
• Migration in order to get a job / improve their work
prospects
• Less often living in a couple than women
Women
• Family at the heart of the migratory project
• A late labour market inclusion
17. - Who is the language training prescribed to ? -
• 20 % of the new migrants are involved
• Up to 400 hours
Not prescribed Prescribed
Women 52 66
Men 48 34
Not prescribed Prescribed
Employed 51 29
Unemployed 24 20
Student 10 7
At home 12 37
Other inactive 3 8
Not prescribed Prescribed
French-speaking Africa
(Maghreb not included)
35 3
Non french-speaking Africa 4 6
Maghreb 39 24
Asia 11 52
Europe and CIS 5 12
Not prescribed Prescribed
Less than 2 years 42 71
Between 2 and 4 years 16 16
Between 5 and 9 years 29 10
10 years and over 12 3
18. - Sub-cohorts building -
Two similar populations apart from the completion of
the language training and a control group :
• Trained cohort
• Counterfactual cohort
• Control cohort
19. ’Trained’ improve
slower than
‘Counterfactuals’ …
... but they increasingly
speak french with their
friends
48
35
62
37
6 4
52
65
38
63
94 96
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2010 2011 2010 2011 2010 2011
Trained Counterfactuals Control
Low Medium and fluent
Level of french by the cohort in 2010 and 2011
Language spoken with friends by the cohort
5 10 16 18
42 42
44
54 42
49
46 51
50
36 42
33
12 8
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
100%
2010 2011 2010 2011 2010 2011
Trained Counterfactuals Control
French French and others Other languages
20. – Conclusion –
• People in distress are identified
• No net impact of the training on the fluency in french
• The training achieves its goal
• However, the beneficiaries are not fully satisfied
• Is the training ambitious enough ?
Prescribed Not prescribed Prescribed Not prescribed Prescribed Not prescribed
Low level 64 36 67 33 56 44
Medium level 21 79 26 74 15 85
Fluent level 2 98 3 97 1 99
All Women Men
In France, there are « official statistical surveys » : like census, labour force survey, …
It means that they have been declared « of public interest » and with a sound methodology by an official body (independant, gathering authorities and civil society).
Internal costs : building the questionnaires, testing them, preparing the data, teaching surveyors, monitoring : about 2 full time equivalents (high skilled statisticians) during 4 ½ years.
NSI national statiscal office
Expliquer CAI
Mêmes personnes réinterrogées + langues et régions
Parcours individuels très différents selon le sexe et la durée de présence en France
+ origine moins important que motif et DP
Hamper
Test de langue au CAI, nombre d’heures proposé suivant le niveau estimé
Involved
Intervertir 52 et 48 pour hommes et femmes
Regarder la colonne ensemble
Comparer bénéficiaires/ non bénéficiaires ?
Ils sont différents et donc difficilement comparables. Tous n’ont pas besoin de la formation linguistique
Variables utilisées : sexe, natio, motif, DP, diplôme, activité, âge arrivée, niveau français, langue durant enfance
Comparer bénéficiaires/ non bénéficiaires ?
Ils sont différents et donc difficilement comparables. Tous n’ont pas besoin de la formation linguistique
Variables utilisées : sexe, natio, motif, DP, diplôme, activité, âge arrivée, niveau français, langue durant enfance