The Long Term Legacy of School Choice
Lead Research Organisation:
Lancaster University
Department Name: Economics
Abstract
This project will evaluate the effect of attending a better school on long term outcomes for secondary school students - such as access to, and success in, higher education, and beyond that to include labour market success reflected in earnings. Unlike other research that looks at the effects of schools attended on outcomes, this research will focus on the effect of NOT going to one's most preferred school. The aim is to focus on the consequences of school choice.
The analysis will be quantitative and will involve detailed statistical analysis of the Longitudinal Study of Young People in England (LSYPE). This dataset follows 17000 young people from age 14 through the education system and into the labour market. They are now 25 and the last sweep of the dataset is about to become available to researchers. The dataset gives rich information about the child, the parents, and the school.
The outcomes that we will go beyond investigate are the educational attainments that is the usual focus of attention for researchers (GCSE's and A-levels) to include well-being, mental health, being bullied, absence, crime, teen sex and pregnancy, and a range of non-cognitive skills (conscientiousness, and the like). Importantly, these students are followed through their higher education experience so we can investigate the impact of being in a second best school on: aspirations to higher education, application, attending an elite institution, dropping-out, and degree class. Finally, the data now contains some earnings and employment information from the subjects' early experiences of the labour market.
School choice involves a trade-off between school proximity and school quality. One might most prefer a local excellent school but if that is not available the choice might be between a local school, that is of satisfactory quality, and a more distant school that is of better quality. The choices that parents make for their children will balance the better school quality against the inconvenience of extra travel time, among other things. This new research will exploit earlier research that analyses the preferences that parents have for schools for their children - based on the preferences expressed by all parents in England for two recent cohorts. This earlier research provides statistical estimates of the extent to which parents are prepared to trade off lesser proximity for better quality.
We use our knowledge of this trade-off in this much richer dataset. We can calculate the distance between home and school attended in the LSYPE data and we know the quality of that school - for example, in terms of % with 5+ A-C GCSEs. And we can see the location of other schools that one might have attended and we observe their quality and proximity. Our estimates allow us to calculate the probability of attending each potential school, from our knowledge of the admission criteria. The work will inform policymakers about the benefits of providing better choices, and will inform parents about the benefits of making better choices.
The analysis will be quantitative and will involve detailed statistical analysis of the Longitudinal Study of Young People in England (LSYPE). This dataset follows 17000 young people from age 14 through the education system and into the labour market. They are now 25 and the last sweep of the dataset is about to become available to researchers. The dataset gives rich information about the child, the parents, and the school.
The outcomes that we will go beyond investigate are the educational attainments that is the usual focus of attention for researchers (GCSE's and A-levels) to include well-being, mental health, being bullied, absence, crime, teen sex and pregnancy, and a range of non-cognitive skills (conscientiousness, and the like). Importantly, these students are followed through their higher education experience so we can investigate the impact of being in a second best school on: aspirations to higher education, application, attending an elite institution, dropping-out, and degree class. Finally, the data now contains some earnings and employment information from the subjects' early experiences of the labour market.
School choice involves a trade-off between school proximity and school quality. One might most prefer a local excellent school but if that is not available the choice might be between a local school, that is of satisfactory quality, and a more distant school that is of better quality. The choices that parents make for their children will balance the better school quality against the inconvenience of extra travel time, among other things. This new research will exploit earlier research that analyses the preferences that parents have for schools for their children - based on the preferences expressed by all parents in England for two recent cohorts. This earlier research provides statistical estimates of the extent to which parents are prepared to trade off lesser proximity for better quality.
We use our knowledge of this trade-off in this much richer dataset. We can calculate the distance between home and school attended in the LSYPE data and we know the quality of that school - for example, in terms of % with 5+ A-C GCSEs. And we can see the location of other schools that one might have attended and we observe their quality and proximity. Our estimates allow us to calculate the probability of attending each potential school, from our knowledge of the admission criteria. The work will inform policymakers about the benefits of providing better choices, and will inform parents about the benefits of making better choices.
Planned Impact
Non-academic beneficiaries include DfE, Local Education Authorities, potential entrants to the "free school" market, headteachers considering the impact of school choice policies, and Regional School Commissioners. There is potential for impact elsewhere - in NI and Scotland for example who are in the process of developing their own databases; and in the US where policy is in limbo right now since Obama's ESSA has had its funds frozen and the new administration has not developed a policy to replace it.
The project will benefit future children and their parents. It will cast light on the importance of the characteristics of schools, not just for academic for achievement, but for the well-being of the child. This will help parents make better choices. We hope to engage the Behavioural Insights Team in the work.
It will also impact on the DfE. It will tell them about how much school quality and school proximity matters for outcomes. It will illuminate, in detail, how the school choice mechanism is working on average - and, importantly, for different types of children..... the gifted and talented, those with SENs, and by gender, ethicity, and month of birth (an indicator of maturity). This will help Whitehall identify choice coldspots and hotspots and the value of moderating them.
As well as central government we hope to impact on the new Regional School Commissioners - whose job is it to go beyond school improvement and consider the coordination of schooling as a whole in their areas.
The project will benefit future children and their parents. It will cast light on the importance of the characteristics of schools, not just for academic for achievement, but for the well-being of the child. This will help parents make better choices. We hope to engage the Behavioural Insights Team in the work.
It will also impact on the DfE. It will tell them about how much school quality and school proximity matters for outcomes. It will illuminate, in detail, how the school choice mechanism is working on average - and, importantly, for different types of children..... the gifted and talented, those with SENs, and by gender, ethicity, and month of birth (an indicator of maturity). This will help Whitehall identify choice coldspots and hotspots and the value of moderating them.
As well as central government we hope to impact on the new Regional School Commissioners - whose job is it to go beyond school improvement and consider the coordination of schooling as a whole in their areas.
Organisations
Publications
Gorman E
(2021)
Heterogeneous effects of missing out on a place at a preferred secondary school in England
in Economics of Education Review
Gorman E
(2021)
Adolescent School Bullying Victimization and Later Life Outcomes
in Oxford Bulletin of Economics and Statistics
Gorman, E
(2021)
Adolescent School Bullying Victimisation and Later Life Outcomes
in Oxford Bulletin of Economics and Statistics
McKendrick, A.
(2022)
Keeping Faith; the the Short and Long Term Impacts of Faith and Faith Schools
Mendolia S
(2023)
COVID-19 vaccination intentions and subsequent uptake: An analysis of the role of marginalisation in society using British longitudinal data.
in Social science & medicine (1982)
Mendolia S
(2018)
Heterogeneous effects of high school peers on educational outcomes
in Oxford Economic Papers
| Description | The project is now completed, following the postponement to facilitate the RA's secondment to DfE. The two major findings are: 1. The current English school choice mechanism that was introducted in 2008 and we find that it works better than the one that it replaced in 2008. We find that that missing out on your preferred school under the new system has no statistically significant adverse effects on any of the child outcomes that we examined, while we find that missiing out on a place at one's prefereed school under the pre 2008 system would have had large negative consequences for: mental health at age 25 - we estimate that the mental ill health score increased by approximately 0.8 (which is large given that the average score is around 2); earnings - we estimate that earnings would be approximately 5% lower if you missed out under the pre 2008 system but NOT under the new system; the probability of attending university would be 14% lower if one missed out under the only system but not under the new one; and the A-level points score, corresponding the grades achieved, would be around 25 points less where the average is close to 400. These are important findings for two reasons. First, the economic analysis (done by Nobel laureate Al Roth of Stanford), that lies behind the way in which the new system works, predicts that it would have the property that missing out would have less serious consequences under this new mechanism. Our work is the first paper that has been able to test this theroetical propostion and we find that the theory is vindicated in practice. Secondly, the change in the mechanism was virtually costless - and the benefits of it seem large. Both the new and old system have, in practice, about 20% of children missing out on their preferred school - but in the old system there were serious adverse consequences of that, while there are none under the new system. 2. We have also explored the mediating effects of bullying - the preferred school is likely to have bee chosen by the parents to be a good match for their child, and so not getting admitted seems likely to imply a worse match and that might manifes itself in being more liley to be bullied. In fact we find only a weak effect of getting into one's preferred school on the probability of being bullied. However, in the course of this work we have explored the effects of being bullied on educational outcomes and other long term outcomes more generally. We find that being bullied (especially bullying that involves violence or a threat of violence) has large adverse outcomes on mental health and income - but a rather small effect on educational outcomes. This is the first research that has explored the long term effects of being bullied during key stage 3 schooling. |
| Exploitation Route | 1. We were granted a 2 hour "policy session" at the RES conference in mid April 2019 where (with other experts in school choice including from the US) we highlighted our findings. Several people from the school choice team at DfE attended and some media engagement occurred. 2. We also organised a 2-day event at Lancaster in september - Al Roth (nobel laureate for his mechanism design work) was the keynote and we invited policy people, the behavioural insights team people concerned, and a few people who are involved in the practical running of school choice mechanisms around the country. We are focussing on getting latter on board with the idea of implementing nudges to improve school choice. |
| Sectors | Education |
| URL | http://humcap.uchicago.edu/RePEc/hka/wpaper/Gorman_Harmon_Mendolia_etal_2019_causal-effects-adolescent-bullying.pdf |
| Description | So far our estimates have been used to secure a contract with the Social Mobility Commission to scope out the effects of changing school admission criteria on social mobility. For example, SMC worries that proximity to good schools plays and important role in reducing social mobility because only richer parents can afford to live near the best schools. They are asking us to explore how it might be possible to simulate what effect alternative admission criteria might have on social mobility. For example, one might replace proximity with a lottery? This work is ongoing and involves merging a model of the determinants of school choice (for example, distance and quality enter into parental preferences ) with the admission criteria to determine who goes to which school. Then we confront this with a model of peer effects to show how the mix of children determines outcomes. We can use this framework to hypothetically reform admission rules and show the consequences, via peer effects, for the children. We have proved ths concept - and will shortly be applying for a lager SMC contract to take the work forward. 4. The research assistant on the project was invited to become a policy intern at the Department for Education where she worked on DfE's new LEO database. The contacts that she made there helped with securing the SMC contract. |
| First Year Of Impact | 2020 |
| Sector | Education |
| Impact Types | Policy & public services |
| Description | Balanced schools - do thay make a difference to social mobility |
| Amount | £30,000 (GBP) |
| Organisation | Government of the UK |
| Department | Social Mobility and Child Poverty Commission |
| Sector | Public |
| Country | United Kingdom |
| Start | 07/2019 |
| End | 04/2020 |
| Description | How simple changes could improve secondary school choices |
| Form Of Engagement Activity | A press release, press conference or response to a media enquiry/interview |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Public/other audiences |
| Results and Impact | Article in "The Conversation" |
| Year(s) Of Engagement Activity | 2020 |
| URL | https://theconversation.com/how-simple-changes-could-improve-secondary-school-choices-141095 |