London schools: the ethnicity debate

23rd February 2015

The TES reported today that a ‘row’ has broken out regarding the role of ethnicity in driving the transformation of London’s secondary schools during the last decade.

In the red corner: a report released today by CfBT Education Trust, following on from their Lessons from London schools report last summer, in which the authors argue that the capital’s pupils now perform better than those elsewhere in the country “regardless of the changing ethnic make-up of the region and regardless of students’ ethnic background.”

In the blue corner: Sam Freedman, who argued in June 2014 that Lessons from London schools underestimated the effect of pupil characteristics (including ethnicity) on the success of the capital’s schools, and Simon Burgess, whose research released in the autumn concluded that the ‘London effect’ is entirely accounted for by London’s ethnic composition.

Somewhere in between these two positions, but arguably closer to the blue corner than the red corner, is analysis by Dave Thomson at FFT. His analysis shows that “around two-thirds of the London effect on mean GCSE grade in 2013 is due to pupil context factors such as ethnicity, free school meal eligibility, special educational needs, first language and mobility.” However, a small ‘London effect’ remains, even when these pupil-level factors, and school-level factors, such as FSM and prior attainment, are controlled for. The remaining difference, although small, may be due to the more traditional, GCSE-based curriculum followed by the capital’s schools – something Burgess also alludes to. Some of the difference may also be accounted for by the school and system-level factors identified by the CfBT’s latest research.

As the various analyses of London’s success continue to trade blows on the significance of ethnicity, two important lessons emerge from the dust cloud:

1.     Whether or not the ‘London effect’ is identified as being due to the capital’s ethnic composition depends a great deal on how we define ethnicity:

·       How many ethnic groups are we including in our analysis, and how are we distinguishing and defining them?

·       We need to be clear whether we’re talking about individual pupils’ ethnic backgrounds, or school or area-level ethnic composition (as was the focus of our analysis for Lessons from London schools, which found that changes in boroughs’ ethnic diversity accounted for around 11% of the change in their GCSE attainment between 2001 and 2011)

·       Talk of ‘ethnic groups’ alone is too crude: we need to also consider the role of migration generation. Some ethnic groups contain predominantly first generation pupils, while others contain many second and third generation pupils. First generation ethnic minorities who arrive late on in their school careers tend not to do as well as their second generation peers, due to language barriers and adjusting to the change in school/curriculum context. When it comes to explaining the London effect specifically, it may be that the capital has more second generation and ‘positively selected’ first generation pupils than other regions, due to the highly skilled labour market and high cost of living there. These are theories which need testing with data, but the fact remains: the generation question is too often overlooked.

2.     There’s a risk that the entire ‘row’ about the effect of ethnicity could become over-simplified: “either London’s success is due to ethnicity or it isn’t.” Taking a balanced view of all the findings, it seems as though pupil characteristics, including ethnic composition, do account for a substantial portion of the London effect. However, the capital’s advantage isn’t completely explained away by pupil characteristics, and explanations are not mutually exclusive. In fact, they could be mutually reinforcing: both an ethnic effect and a system/curriculum-based effect could be combining to make an extra difference for ethnic minorities in London compared to elsewhere.