BILINGUAL CHILDREN DO BETTER IN TESTS: US evidence that speaking two languages in the early years gives kids a head start in life

Parents who worry about confusing their babies by speaking two languages at home don''t need to fret: children between the ages of 3 and 12 who speak two languages do better in cognitive tests than similar kids who only speak one language. That is the conclusion of research by Ainhoa Aparicio Fenoll, to be presented at the Royal Economic Society''s annual conference at the University of Bristol in April 2017.

By analysing the test scores of American children who were born to Latino immigrants and spoke both English and Spanish, and comparing them with the scores of similar children who spoke only English or only Spanish, the study shows that learning two languages has not held the children back – in fact, the opposite. In tests designed to test their cognitive skills, they did better, even when their background, schooling or wealth was taken into account.

''Parents deciding on which language to speak to their children often face contradictory recommendations,'' the author says, adding that this new evidence suggests that whatever the root cause of the difference in educational attainment for Latinos in the United States, the need to learn two languages in early years isn''t to blame.


This paper tests whether bilingual children have better cognitive skills than comparable monolingual children. A high and increasing number of studies analyse differences between bilingual and monolingual children, focusing on specific aspects of the cognitive process and using behavioural tests performed on a group of voluntary subjects.

My study presents three novelties:

• it uses a big and representative sample of children born to Latino immigrants living in the United States;
• it focuses on cognitive standardised test scores that measure a wide variety of aspects and are widely used in practice to evaluate students;
• and it compares monolinguals and bilinguals, taking account not only of demographic and socioeconomic characteristics but also home and school inputs.

The results show that bilingual children outperform monolinguals taking the test in their one language.

More than half of the world is able to speak two or more languages. According to the American Community Survey, 20% of US residents in 2012 were bilingual. Bilingualism is considered to shape bilinguals'' brains and this aspect of bilingualism has captured most of researchers'' attention.

In particular, there has been a long lasting and still very intense debate on the consequences of bilingualism for children''s cognitive skills. These skills determine their future education attainment and labour market performance. Hence, if bilingualism improves cognitive skills, this could be behind the evidence of a positive association between bilingualism and academic outcomes in the sociology and education literatures and behind the positive effects of home language proficiency on labour market outcomes and social integration of immigrants.

On the other hand, a negative impact of bilingualism on cognitive skills could explain the native-immigrant gap in cognitive skills.

Experts are far from reaching a long-lasting agreement on whether bilingualism is positive or negative for children cognitive skills. Before educators and practitioners could adopt the new positive view that is widespread among academics, this view has been challenged by a number of recent papers. In line with experts'' disagreement and probably as a consequence of it, parents deciding on which language to speak to their children often face contradictory recommendations.

In this paper, I estimate the magnitude and statistical significance of cognitive differences between bilingual and monolingual children as measured by standardised cognitive tests with consequences for real life outcomes.

For my estimation, I make use of information on a representative sample of children born to Latino immigrants living in the United States provided by the New Immigrant Survey (NIS).

Crucially for my analysis, this survey provides information on language proficiency and cognitive test scores obtained from standardised Woodcock tests administered to children aged 3 to 12. A set of variables on English and Spanish proficiency (speaking, understanding, reading and writing skills) is used to classify children as bilinguals, Spanish monolinguals or English monolinguals.

In contrast with previous literature focusing on specific tasks, standardised testing is a relevant outcome per se because these tests are an integral component of academic progress and success and are often used to assign students to different tracks.

The NIS also includes a wide set of variables that allows me to make the bilingual-monolingual comparison of children who are similar in many dimensions: family and individual characteristics, parental labour market outcomes, home and school inputs and parenting behaviours.

Finally, the design of the NIS includes an experiment in which the language of the test is assigned randomly. I use this feature to compare bilinguals to Spanish monolinguals in tests in Spanish and bilinguals to English monolinguals in tests in English. In both cases, I find representative evidence on the bilingual advantage.

Twinkle Twinkle Little Estrella: Representative Evidence on the Relevant Bilingual Advantage – Ainhoa Aparicio Fenoll, Collegio Carlo Alberto, IZA and CHILD, Via Real Collegio 30, Moncalieri (TO) 10024, Italy