skip to content

Keynes Fund


Summary of Project Results

In this project we studied how children are allocated to schools in the English school system, with the aim of modelling parental preferences for school characteristics (e.g. school quality) and the constraints faced by families (the schools that their children can actually access). The latter is determined by the admissions criteria used by any school that they might want to attend.

We know that an individual’s permanent income depends strongly on their accumulated human capital. One long-term approach to reducing income inequality is therefore to focus on inequalities in schooling. Undoubtedly, part of this inequality arises from differences in family background, but schools also matter and schools differ. School performance data show that pupils make markedly different levels of progress through secondary school and that the socio-economic gap in pupil achievement widens due to the different schools attended by different pupils. This means that which school a child attends will, on average, matter for their achievement. School assignment in England is determined by a system of school choice: since 1988, parents have had the right to nominate which schools they would like their child to attend. These nominations are combined with school admissions policies to determine the final allocation of pupils to schools. In this project, we collected new data on schools' admissions criteria to determine children’s chance of access to high quality schools. Future work will combine it with existing data on parental preferences to explore the trade-offs that parents make when choosing schools.

Our research ultimately seeks to provide evidence for England on:


  1. What characteristics of secondary schools are valued by parents? What is the strength of the preference for academic excellence? What is the trade-off that parents make between distance from a school and academic quality for secondary schools?

  2. Do parents from different socioeconomic backgrounds have different preference parameters, and what weight is given by different groups to academic quality?


To answer these questions, one needs granular data on both parental preferences and school admissions criteria. We had already secured access to administrative data on parental preferences and the characteristics of pupils in each school. Specifically, we used the National Pupil Database, an administrative dataset covering all pupils in the state sector in England, including demographic data on pupils, their exam score history, and their home location and schools attended. We linked that to national data on parental preferences for schools, specifically all submitted preferences for schools by every parent seeking a place at a state secondary school in England. These data record the actual preferences in rank order submitted by parents on the official forms. As such, this is an improvement on the parental recall of submitted preferences that we used in our previous analysis of primary school choices (Burgess et al., 2015).


School admissions

There is no national data on the admissions criteria used by schools. Hence the first, and most important, objective of this project was to undertake a significant amount of primary data collection on school admission criteria. Unfortunately, COVID struck part way through the grant, which severely delayed this data collection. Despite the significant hurdles we faced, we completed the collection of primary data on the full admissions criteria of almost every state-funded secondary school in England (3,248 schools). This information provides an unparalleled level of detail on each school’s admissions criteria. For example, for each school we know each separate admissions priority in order (for example, looked after children receive the highest priority, followed by siblings of children at the school, followed by those inside the catchment area, followed by those outside the catchment area by straight line distance). When a catchment area is mentioned, we have recorded the precise details, including the spatial coordinates of the catchment boundaries. Where a feeder school(s) is mentioned we have recorded the school IDs.

These data allow us to re-create much more accurately the sets of schools that parents are making their choices from. This is essential to estimate parents’ preferences for schools from their submitted lists of schools. Without such data, constraints on choices (such as some desirable schools being unavailable), may be mistaken for preferences (for example, not valuing high academic performance).

With the data collection complete, we have drafted a working paper which we are currently finalising for submission to a journal. Descriptive analysis suggests:


  • Admissions criteria vary dramatically in number and complexity by school, and obtaining accurate information on criteria is difficult, even for a researcher. With the introduction of quasi-autonomous Academy schools that can set their own admissions criteria (bounded by a national admissions code of practice), there is far more complexity in the system from the perspective of a parent trying to make a choice. This makes it difficult for parents to determine whether or not their child can access a particular school, which will tend to undermine the school choice system.

  • Different school types vary in the number and complexity of the criteria they use. Religious schools use a larger number of often very complex criteria;

  • Using Ofsted ratings as one indicator of quality, we showed that de facto access to good/outstanding schools varies by postcode. Students in the poorest neighbourhoods have a lower probability of accessing a higher performing school, even after controlling for a range of individual demographic and locality factors.


As well as a working paper, an early think piece on school choice was presented to the Sutton Trust, written up as a report and then showcased in the Sutton Trust’s work on this issue (links below).


CI Simon Burgess wrote an influential blog about the way that we could make the school choice process fairer, citing data from this project.


Members of the project team (Burgess, Greaves) have submitted a grant application to build on this initial work.



Burgess, S., Greaves, E., Vignoles, A. and Wilson, D., 2015. What parents want: School preferences and school choice. The Economic Journal, 125(587), pp.1262-1289.

Burgess, S., Greaves, E. and Vignoles, A., 2019. School choice in England: evidence from national administrative data. Oxford Review of Education, 45(5), pp.690-710.



Prof. Anna Vignoles


Professor Anna Vignoles is a Professor of Education at the Faculty of Education, University of Cambridge, a Fellow of the British Academy, a trustee of the Nuffield Foundation and a member of the ESRC Council. Her research interests are Economics of Education, Equity in Education, Economic Value of Education, Widening Participation in Higher Education, Quantitative Methods.


Share this Project