Do resources matter? This question remains controversial in the economics of education as many studies find no relationship between school resources and educational outcomes. Yet, improving educational performance, especially of ‘hard to reach’ children, is a key area for government policy. This is particularly the case in countries like the UK and the US where the fact that many adults have poor basic skills is frequently attributed to people being ‘let down’ by the education system. In the UK, particular attention has been paid to schools in inner-city areas, where many pupils face problems of socio-economic deprivation and where there has been a concern about educational underachievement for many years. We study a flagship policy of the UK government – the Excellence in Cities (EiC) programme – which has been designed to address these problems. This enables us to consider not only whether this particular policy was successful, but also to contribute to the more general debate as to whether resources matter and in what circumstances. We look at the distributional impacts of resources in a way that is not addressed by much of the existing literature. Excellence in Cities was launched in 1999 in over 400 secondary schools in England and since then progressively increased in coverage. It is now implemented in about a third of all secondary schools (over 1,000 schools). The policy involves allocating resources to Partnerships of Local Education Authorities (LEAs) and LEA maintained schools within their respective regions. Since there is considerable heterogeneity in the degree of disadvantage and school performance within LEAs, the policy does not cover every disadvantaged or poorly performing school in England and hence it is possible to find a comparison group of schools outside the policy. The three core strands of EiC are as follows: Learning Mentors, to help students overcome educational or behaviour problems; Learning Support Units, to provide short-term teaching and support programmes for difficult students; and a Gifted and Talented programme, to provide extra support for 5-10 per cent of pupils in each school. Other aspects of the EiC policy are the designation of particular schools as Specialist (i.e. in particular subjects) or Beacon (to disseminate good practice). Schools which are successful in their application for such status receive significant amounts of money. Finally, EiC also involves the creation of City Learning Centres (to provide ICT facilities) and Education Action Zones (where there is an emphasis on the sharing of good practice). We evaluate the average impact of EiC on educational attainment and attendance at school over time since its introduction in 1999. Thus, we assess the extent to which the whole range of activities carried out as a result of EiC funding led to an improvement in important educational outcomes. We focus on pupil-level attainment at age 14 (the end of Key Stage 3) and a measure of school attendance (the percentage of half days missed). We consider variation in the effect of the policy according to when it was introduced to different areas; time since its introduction; the level of disadvantage of the school (and of pupils of different abilities within these schools). Our methodology is based on a comparison between outcomes in schools where the EiC has been in place and schools in an appropriate comparison group before and after the policy was introduced. Specifically, we use a difference-in-differences approach that is combined with statistical matching. Our results show a positive effect of EiC on pupil attainment in Mathematics (though not in English) and on school attendance. A simple assessment suggests that the policy is cost-effective – at least for more recent years. However, the effect of the policy is heterogeneous along a number of dimensions: it is stronger the longer the policy has been in place; for disadvantaged schools; for medium-high ability pupils within disadvantaged schools. Our results show that ‘resources matter’ but that it is difficult to help the ‘hard-to-reach’ using this level of resources (i.e. low ability children within disadvantaged schools). For such children, different and probably more intensive policy treatments may be required – ideally earlier in their schooling career.
