The number of primary school pupils in England who do not speak English as a first language has been growing in recent years. Sandra McNally examines whether this is damaging the educational outcomes of native English speakers, concluding that this should not in fact be a cause for concern.
The number of non-native speakers of English in primary schools in England has increased by a third over the past 10 years. Now, roughly one in nine children between the ages of 5 and 11 do not speak English as a first language. A significant driver of this change has been immigration, though the trend has also been influenced by higher birth rates among ethnic minority groups.
The change has led some commentators to fear that the impact might be detrimental to the educational attainment of native English speakers. They fear that teachers’ time will be taken up giving extra help to children who do not speak English as a first language.
Yet several studies have shown that first and second generation immigrants are, on average, better educated than the native population. This suggests that there might be things about the children of immigrants – such as having better educated parents – that can compensate for any lack of language fluency at an early age. In that case, native English speakers would not necessarily suffer from having such children as their peers.
Our research analyses a census of all children in English schools – the National Pupil Database – to explore these issues. We look at whether there is an association between the proportion of non-native English speakers in a year group and the educational attainment of native English speakers at the end of primary school – and whether it can be interpreted as a causal relationship.
We also split the data into white and non-white non-native speakers to see if there might be different effects. Although the latter group is more important numerically, the former has grown very sharply after the eastern enlargement of the European Union (EU) in 2005.
We find that there is negative (although modest) raw correlation between the educational attainment of native English speakers and the proportion of non-native speakers in their year group. This correlation falls by at least a half if we control for just a few demographic characteristics of native English speakers. It disappears altogether if we control for the schools attended by non-native English speakers – and it is little affected by successively adding more detailed controls.
These results indicate that the negative correlation in the raw data is an artefact of the schools attended by non-native English speakers. For example, they typically attend schools with more disadvantaged native speakers. Taking account of the fact that non-native English speakers are more likely to attend schools with less favourable characteristics, there is zero association between their presence in greater numbers and the educational attainment of their native English-speaking peers.
This result is also true if we look at younger cohorts (age 7 instead of age 11) and if we look at the number of languages spoken in the year group instead of the percentage of non-native English speakers. We explore many different aspects of heterogeneity, for example, looking at native English speakers who are disadvantaged, who are of low ability and who are based in London.
We also divide non-native English speakers into those who appeared in the school census in the last two years of primary school versus those who were in the school census before that time. This affects the raw association between the percentage of non-native English speakers and the educational attainment of native English speakers. But once demographics of native speakers and school controls are added, the effects go to zero in almost every case.
Under certain assumptions, our estimates can be interpreted as reflecting a causal relationship. While we cannot fully test these assumptions, our analysis strongly suggests that negative causal effects of non-native English speakers on the educational attainment of native English speakers can be ruled out.
We also use another research strategy to look at the relationship between the percentage of white non-native English speakers and the educational attainment of native English speakers. This strategy uses the fact that the number of white non-native English speakers grew dramatically after the EU’s eastern enlargement in 2005.
Since many of the new immigrants were Polish (and likely to be Catholic), there was a big rise in the demand for Catholic schooling. The data show a much larger increase in the percentage of white non-native English speakers in (state) Catholic schools after 2005 compared with other schools.
We use this as a ‘natural experiment’ to see if there were consequences for the relative educational attainment of native English speakers in Catholic schools. The results for reading and writing show no clear impact, but there is some evidence for a small, positive effect in the case of maths. In other words, native English speakers at Catholic schools that saw a strong relative increase in white non-native speakers benefited to a small extent in their maths results.
We can only speculate as to the possible reasons for this result. It may be the fact that immigrants from East European countries are better educated and more attached to the labour market than the native population. The children of such immigrants may be a welcome influence in the schools they attend.
The two different research strategies apply to different populations. The first shows associations that are applicable to all schoolchildren. The second – making use of eastern enlargement – only estimates the effects on native English speakers in Catholic schools who were exposed to an increase in white non-native speakers after enlargement. Thus, the latter results cannot be extrapolated to other contexts.
But both strategies suggest that negative effects of non-native English speakers can be ruled out. Thus, the growing proportion of non-native English speakers in primary schools should not be a cause for concern: this trend is not detrimental to the educational attainment of native English speakers.
This article summarises ‘Non-native Speakers of English in the Classroom: What are the Effects on Pupil Performance?’ by Charlotte Geay, Sandra McNally and Shqiponja Telhaj, Centre for the Study of the Economics of Education Discussion Paper No. 137 published March 2012.
Please read our comments policy before posting.
Note: This article gives the views of the author, and not the position of the British Politics and Policy blog, nor of the London School of Economics.
Sandra McNally is director of CEP’s research programme on education and skills (http://cep.lse.ac.uk/_new/research/education/default.asp).
Charlotte Geay is at the Paris Graduate School of Economics, Statistics and Finance.
Shqiponja Telhaj is a lecturer at the University of Sussex and a research associate in CEP’s education and skills programme.