RESEARCH

The Effects of Youth Clubs on Education and Crime 

IFS Working Paper 24/51

Youth clubs are community-based after-school programmes typically offered free of charge to teenagers in underprivileged neighbourhoods. I provide the first causal estimates of their effects on education and crime, leveraging quasi-experimental variation from austerity-related cuts, which led to the closure of 30% of youth clubs in London between 2010 and 2019. I use difference-in-differences research designs and novel data to compare neighbourhoods affected by closures with those unaffected. Teenagers in areas affected performed nearly 4% worse in national high-school exams. Youths aged 10 to 17 became 14% more likely to commit crimes. Youth clubs provide key support in a lasting manner, particularly to teenagers from low-income backgrounds. The effects are due to youth clubs offering unique amenities that support positive behaviours rather than mere incapacitation. Closing youth clubs was not cost-effective; for every £1 saved from closures, there are associated losses of nearly £3 due to forgone returns to education and crime costs.

Runner up for Best Paper at Royal Economic Society PhD Conference 2023 

Selected for the EALE Tour 2025

Selected media coverage: VoxEU, The Guardian (in 2024), BBC News, Evening Standard, The Independent, Economics Observatory, The Guardian (in 2025)

Other impact: National Youth Agency, Schoolsweek, UK Parliament-Hansard,  UK Parliament TV (minutes 10.03.15, and 10.33.50), IfG 

https://drive.google.com/file/d/1dQvweis-NIk959t3SVWh2MnOyNf2-38k/view?usp=sharing

Specialised Courts and the Reporting of Intimate Partner Violence: Evidence from Spain 

with Jorge Garcia-Hombrados and Marta Martinez-Matute - Journal of Public Economics  (2024)

This paper assesses the effect of the creation of specialised intimate partner violence (IPV) courts on the reporting of IPV, and the incidence of IPV homicides in Spain. We find that the opening of a specialised IPV court increases the reporting of IPV by nearly 122 offences per 100,000 inhabitants, or 28% in the preferred specification. The rise in reporting is primarily driven by an increase in the reporting of moderate offences. We do not find conclusive evidence on the effects of specialised courts on IPV homicides.

Selected media coverage: El Diario, ABC 

Minimum Legal Drinking Age and Educational Outcomes 

CEPR DP19733, IZA DP 17507

with Manuel Bagues - submitted

Over the past decades, many European countries have raised the minimum legal drinking age (MLDA) from 16 to 18 years. This study provides novel evidence of the impact of this policy on educational outcomes by exploiting the staggered timing of MLDA changes across Spanish regions. Raising the MLDA decreased alcohol consumption among adolescents aged 14–17 by 8 to 18% and improved their exam performance by 4% of a standard deviation. This effect appears driven by alcohol's direct impact on cognitive ability, as we find no significant changes in potential mediators like use of other substances or time spent on leisure activities, including socialising, sports, gaming, or internet use. We also observe a decrease in tranquilliser and sleeping pill use, suggesting improved mental health. Our findings indicate that reducing teenage alcohol consumption represents a significant opportunity to improve educational outcomes in Europe, where youth drinking rates remain notably high.

Selected media coverage: NadaEsGratis

Gangs of London: Public Housing, Bombs, and Knives 

with Richard Disney, Tom Kirchmaier and Stephen Machin

We analyse the spatial distribution of local street gangs operating in London from 1990 to 2015, focusing on how housing characteristics determine gang presence. High-rise public housing estates built post-World War II are more likely to become gang turfs than areas with low-rise or no social housing. To resolve any potential reverse causality between the socio-economic characteristics of gang areas and the presence of public housing, for instance, if high-rise social housing was constructed in areas with higher criminality, the London bombing Blitz of 1940-41 is utilised as a shock to urban development. Bomb damage led to the construction of high-rise post-war public housing and, therefore, the formation of gangs in the later period. We then show that gang presence is correlated with a higher incidence of knife crimes with injury and a higher incidence of youth crimes. 

Selected media coverage: BBC World News (TV), LSE IQ, The Economist, Nada Es Gratis

Other Impact: Behavioural Insights, UK Parliament, Mayor of London, Action on Armed Violence

Money for nothing? Short- and long-run effects of paying disadvantaged teenagers in full-time education

with Jack Britton, Nick Ridpath, and Ben Waltmann

We evaluate the short- and long-run effects of the largest conditional cash transfer program implemented in a high-income country. The Education Maintenance Allowance provided low-income teenagers with up to $70 per week (in 2024 US dollars) to remain in full-time education beyond the compulsory leaving age. Exploiting the programme’s staggered rollout across local areas in England, we find that participation in full-time education increased by only 2 percentage points among the poorest students, substantially less than suggested by evaluations of a pilot phase of the programme. The programme likely reduced part-time work among enrolled students and may have lowered crime. We find no improvements in test scores, no effect on qualifications beyond the lowest level, and a negative effect of around 2-3 percent on the earnings of eligible young people in their twenties. While reductions in crime may have generated some social benefits, these are small relative to the programme’s substantial costs. Our results suggest that conditional cash transfers may be substantially less effective in settings with well-developed safety nets for families with children in education.