Tuesday 23 November 2021

The Effects of College Capital Projects on Student Outcomes

 Stephen Gibbons, Claudia Hupkau, Sandra McNally, Henry Overman

About half of school leavers in England attend colleges of Further Education (FE), though these colleges are often considered the poor relation of schools and universities, enrolling lower achieving students and spending less per student (Britton et al. 2019). Capital expenditure accounts for about 10 per cent of FE College expenditure and in the 2020 budget the UK government committed £1.5 billion over five years to bring college facilities up to a good level.

Will this investment make much of a difference to student outcomes? In a Centre for Vocational Education Research (CVER) paper published today, we suggest that it will help - getting more students to a good upper secondary qualification, increasing enrolment in higher-education and even improving employment outcomes.

Our analysis uses information on capital expenditure programmes undertaken between 2006 and 2009, linking this to administrative data on individual educational and labour market outcomes up to 2017 (using the DfE Longitudinal Educational Outcomes data). We focus on large capital projects only. We find that these projects take about three years to complete and that changes in student outcomes take place at that time or the year after. We find that large capital grants increase the share of students enrolled on upper secondary courses that lead to “good” qualifications such as A-levels or BTECs (i.e. at Level 3). This matters because less than half of young learners entering FE colleges progress to these courses (Hupkau et al. 2016).  Level 3 qualifications are associated with higher earnings (McIntosh and Morris, 2016) and are a pre-requisite for university. Conditional on enrolment, large capital grants do not affect achievement. This is a still a good outcome because it shows that enrolments go up with no (negative) effect on achievement rates. Investment in capital projects also affects the probability of enrolling in higher education. The magnitude of effect is within the same ballpark as Machin et al. (2020) who consider the effect of marginally achieving a good grade in GCSE English on enrolment to upper secondary education and university degrees. Furthermore, any benefits from capital projects affect multiple cohorts of students.

These effects persist even after we allow for the fact that FE colleges see a marked change in student composition after the completion of capital projects – they attract students with higher prior achievement and a higher proportion of “non-poor” students (i.e. who did not receive free school meals when in school) – although there is no overall increase in the number of students. There is also a higher probability that students will achieve sustained employment. Effects are usually larger for the largest grants.

There are several reasons why capital expenditure may have these effects. First, substantial capital expenditure on new equipment, laboratories or workshops may improve learning on courses that rely on specific and costly assets (for instance, engineering). Second, better buildings may improve the learning experience. Safe, clean, and appealing learning environments – with no overcrowding, good lighting and heating - could improve concentration and lead to greater student and teacher morale and effort.  On the other hand, large capital expenditure projects may be disruptive and positive effects may take time to materialise. As we show, these positive effects also partly reflect changes in the composition of student intake – improving outcomes for colleges that receive grants but not necessarily for the whole system.

Our study takes advantage of two features of the data and the way the capital expenditure program was implemented to better get at the causal impact of expenditure on outcomes. First, we use rich data on student-level outcomes and characteristics, which allows us to show that improved outcomes don’t simply reflect improved intakes. Second, we show that improved outcomes aren’t explained by expenditure being targeted at colleges that were already improving. Results improve at investing colleges even when compared to a control group of colleges that will benefit from investment in the near future.

Our study is one of very few to evaluate the effect of capital expenditure for students in post-secondary education. Most academic studies evaluate effects in schools, mostly in the US. Our use of micro-data, and our careful attention to causality – allow us to go beyond the government’s own analysis of FE capital expenditure (Business Innovation and Skills, 2012) – which only found small effects on student numbers and no effect on achievement or retention. Our study suggests that these effects were under-estimated.

Our results show that capital investment in college infrastructure has a visible effect on student outcomes within a reasonable timeframe. Investing in capital infrastructure can benefit many cohorts of students and is best considered a long-term investment. These results are, however, reassuring for policy makers who may be more concerned about short-term returns as they show that for large capital projects, the benefits materialise as soon as the project is complete.