Key Findings of the OECD Policy Review of Migrant Education

and

Implications for Language Education Policy for Immigrant Children

Miho TagumaProject Manager of Policy Review of Migrant Education

Intergovernmental Policy Forum2-4 November 2010 1

Stock of foreign-born, as percentage in population

Brief Background

2

Project’s overarching question

What policies will promote successful education outcomes of first and second generation immigrant students?

• Focus:

– Education outcomes (Student Performance, Participation, Access) vs labour market outcomes

– Education policies vs immigration, housing, social and labour market policies

• Scope: Pre-school, primary and secondary education

• Review countries (6 countries): Austria, Denmark, Ireland, the Netherlands, Norway, and Sweden

• Background reports (above + 6 countries): Finland, Hungary, Korea, Spain, Turkey, United Kingdom

• Working methods: Mix of quantitative and qualitative

– Desk-based research: i) statistical analysis for 6 countries +: national statistics, PISA, PARLS and ii) literature reviews, and country background reports to establish facts and explore factors that are linked to education outcomes, and consider policy implications

– Fact-finding and policy review visits to consolidate the facts and suggest effective policy options for the countries concerned 3

• PISA shows: At age 15, marked performance differences in reading are observed.

400

450

500

550

600

Score Native students -

First generation immigrant students OECD average performance in reading

38 pts

Roughly equivalent to one year of

schooling

(science -proxy)

4

Second generation immigrant students

1 How do immigrant students perform?

Student-level factors

Accounting for students' socio - economic backgroundAccounting for students' socio - economic background and language spoken at home

Performance difference in reading

-100

-80

-60

-40

-20

0

20Score point difference

• “SES” and “speaking a different language at home” largely explain the performance gap between the two groups in many countries. But they are not the only reasons.

2 What can explain the gap/ non-gap?

38 pts

Roughly equivalent to one year of

schooling

(science -proxy)

• Other factors: availability of educational resources at home, reading at home at a young age, and participating in ECEC, etc.

5

School-level factors

A few examples….. More hours per week for language learning in regular class

 Australia ++Austria - - -Belgium ++

Flemish Com. ++++Canada +Denmark ~~Germany ~~Greece ++Ireland ~~Italy +++Luxembourg +++++

Netherlands ~~

New Zealand ++++Norway +++Portugal ++++Spain +++Sweden ~~

Switzerland ++

United Kingdom ~~ ~~ Relationship with performance is not statistically significant- / + Less than 20 score point change in reading performance- - / ++ Between 21 and 40 score point change in reading performance- - - / +++ Between 41 and 60 score point change in reading performance- - - - / ++++ Between 61 and 80 score point change in reading performance- - - - - / +++++ More than 80 score point change in reading performanceSource: OECD PISA 2006 database

6

Policy target and policy alignment Government Steering Tools

Ministry of Education

Other ministries

Regional governments

Students (Primary Target)

Parents

Teachers/ school masters

Teacher education univ & colleges

Communities