The impact of English free schools

Introduction

Within the US, there are currently debates over "school choice" involving charter schools, which are privately run schools funded by public money, and voucher programs, which allow parents to use public money to pay for private schools. Proponents of charter schools and voucher programs argue that they offer more choices for parents and produce better academic outcomes for students. Those who argue against these programs say that they devalue public education and may cause segregation between students of varying racial and socioeconomic backgrounds.

Similar debates are occurring in England, where proponents of what are called “free schools” advocate for privately-run schools to be funded with government money, with some arguing that increased competition between schools will cause overall improvement in all schools. Similarly to the US, however, there are concerns from opponents of free schools that such schools could impact negatively on existing schools and increase social segregation.

These debates were at the heart of new research from the University College London Institute of Education, which focused on the mainstream free school population as a whole and the average effect of free schools on their neighbors. The aims of this research were threefold:

1. To analyze the potential impacts of free schools on their neighbors and whether neighbors perceived any competition and responded by taking any action.
2. To analyze whether the opening of free schools was associated with any improvement or deterioration in student attainment in neighboring schools
3. To inquire whether the opening of a free school was associated with any increases or decreases in social segregation in the surrounding local area.

The approach to analyzing potential effects on student outcomes was based on matched difference-in-differences and fixed effect panel estimators, drawing data from the National Pupil Database. The research developed a survey of neighboring schools and case studies of localities in which a free school opened. A ‘neighbor’ was identified when a school experienced the opening of a free school of the same phase in its own neighborhood. A neighborhood was defined as the travel distance to a school’s ninth-nearest school, with evidence presented for the validity of this definition. Using these definitions, over 10% of mainstream primaries and 35% of secondaries were shown to have become neighbors of a free school by 2020.

Results

The research found that free schools were more ethnically diverse than their neighbors, but less diverse in representation of students of lower socioeconomic status and those with special educational needs. Additionally, free schools were not high performing. Primary free schools performed on average worse than a matched sample of schools with similar demographics. They also performed worse on average than their neighbors. Secondary free schools performed on average no better or worse in the main General Certificate of Secondary Education (GCSE) attainment and progress measures than a matched sample or their neighbors. 

Analyzing choice and competition, the data showed that free schools did affect student enrollment in neighboring schools, which declined on average. Competition, rather than collaboration, was the dominant form of perceived relationship with the nearest free school among surveyed neighbors. Perceived competition, particularly over student recruitment and popularity among parents, was stronger where the free school was seen by respondents to appeal to advantaged students or where there were surplus places. Schools competing with a free school were likely to also report adverse impacts on their school, including to student enrollment, associated funding, and popularity among parents.

There was a strong relationship between perceived competition and reported action taking among surveyed neighbor schools. The strongest association was to ‘externally-focused’ actions, which included marketing, promotion and extra-curricular activities. Competition was also associated with ‘accountability-focused’ actions, including placing more emphasis on core curriculum subjects, student attainment in exams and inspections grades. Competition did not however predict ‘internally-focused’ actions, relating directly to the quality of teaching and learning. This points to free school competition spurring schools to deploy more resources to improve their appeal or performance in external metrics, in a race to recruit from a fixed pool of students, but without a focus on improving classroom practices.

Case study schools with surplus places also reported cutting staffing and curriculum. Several heads of these neighbor schools predicted their “solidly good school” would be shut. As these schools tended to serve deprived context, students would not only be dispersed, but a deprived local community would lose the school as a community resource while the local free school orientated to more middle-class communities.

Analyzing student attainment in neighboring schools, the data showed no improvement or deterioration on average in student attainment in English and Maths in primary schools. In secondary neighbors, there was a modest increase in student attainment in English and Maths. There was also evidence, however, that social selection may have been a mediator of free school competition translating into this improvement. Improvement occurred among secondary neighbors that experienced, after a free school opened, a large increase in their percentage of students who had high prior attainment and were not eligible for Free School Meals (FSMs) or were not White British (WBri). These schools already served more advantaged intakes on average prior to a free school opening. By contrast, there was little or no improvement among secondary neighbors experiencing a large increase in students who had low prior attainment and were either eligible for FSMs or were WBri. These schools already served more disadvantaged intakes prior to a free school opening.

Analyzing social segregation, primary free schools were found to be associated with a modest increase in segregation on average. While the trend in England was toward decreasing segregation, areas in which primary free schools opened saw an opposite trend. This related to ethnicity, with students less likely to meet peers from other ethnic backgrounds at school after the primary free school opened. In the secondary phase, increased segregation was not statistically significant on average, but there was a modest increase in segregation for White British students in areas with lower ethnic diversity and rural areas. Drawing on case study data, the research attributed increased social segregation to selective competition between schools and to different ways in which some free schools created new options for parents to choose schools that were more homogenous than the local area. This included both “self-segregation” by minority ethnic parents and perceived “white flight”.

Conclusion

While free schools are a diverse population, they have had observable impacts on their neighbors. Neighbors on average lost students, commonly perceived competition and, where they did, were likely to respond to a free school’s presence. This provides some support to policy claims that free schools force existing schools to take new action, but the research also found clear disruptions to such claims. Neighbors rarely saw free school competition to concern attainment or innovation, rather relating it to student numbers, funding and selective competition over students by socioeconomic status. Free school competition spurred neighbors to deploy more resources to improve their appeal and external quality metrics, but typically without a related focus on classroom practices. While there was a modest estimated increase in student attainment in English and Maths in secondary neighbors on average, there was also evidence to suggest social selection was a mediator of free school competition translating into improvement in secondary schools. Social segregation increased where primary free schools opened, particularly by student ethnicity.

This brief is based on the following published report: Higham, R., Anders, J., Chouhy, G., Green, F., Henseke, G., & McGinity, R. (2024). The Free Schools Experiment: Analysing the impacts of English free schools on neighboring schools. IOE, UCL's Faculty of Education and Society.