Child migrants with and without parents: census-based estimates of scale and characteristics in Argentina, Chile and South Africa

This paper analyses children’s internal and international migration, using individual-level census data, in the three middle-income countries of Argentina, Chile and South Africa. It defines child migrants as under 18 year olds whose usual residence was in a different country or province five years prior to census. Specifically, this paper examines three issues:

• types of migration involving children of different ages (internal/international)
• adults with whom migrant children reside at destinations (family/ non-family)
• children’s schooling levels, work and housing at destinations. 

Correlations between these three are analysed, with the idea that where children migrate to, the people they live with and their characteristics at destinations are related issues.

The paper finds that:

• around four per cent of children were international or internal migrants, involving 1.4 million children and representing a quarter of all migrants
• definitions affect age-profiles. Migration defined by birthplace rather than residence estimates a lower involvement of children, but not by much – the big difference is between migrant stocks and flows
• a conservative estimate suggests that in the three countries over seven per cent of children (migrant and non-migrant) resided independently of adult parents or siblings
• in South Africa where data was available, just four per cent of independent children had both parents dead. Over 10 thousand were international migrants, and 112 thousand internal migrants. This represented nine per cent of child migrant flows
  
• independent child migrants had worse shelter at destinations, and this contrasts with dependent child migrants who seemed not much different from non-migrants in their type of shelter
• average schooling was around six years for independent child migrants, and whilst similar between internal and international migrants, this was nearly two years more than dependent migrant children.

The paper concludes that evidence suggests that there is significant scope to build on available data and develop a more accurate, coherent and useful understanding of child migrants and their role in the development process.