Monday, 9 October 2017

What are the labour market outcomes associated with vocational education and training?

Pietro Patrignani is a Senior Economic Consultant at London Economics working on CVER projects. In this blog he looks at the outcomes for those on vocational paths using newly matched data.

What’s new?

For the first time, the matched Longitudinal Education Outcomes (LEO) data has been made available for analysis of qualification attainment and labour market outcomes in England. This dataset combines information from different school (National Pupil Database), Further Education (Individualised Learner Record) and Higher Education (HESA) data sources in England with labour market outcomes information from HM Revenue and Customs and the Department for Work and Pensions. The key advantage of this matched dataset is that it contains detailed information from administrative data sources on both labour market outcomes and also early scholastic attainment (including Key Stage 1 and Key Stage 2 test score information), but is not restricted to a small sample of individuals (as in the 1970 British Cohort Study).

What did we do?

The study [1] has focused on the cohorts undertaking Key Stage 4 schooling in 2001/02, 2002/03 and 2003/04 (aged 15 at the start of each respective academic year) and assessed the association of vocational attainment with labour market outcomes at age 26 (tax years 2012/13 to 2014/15). To undertake the analysis, on one side, the different educational datasets were combined to generate the highest qualification (vocational or academic) attained by each learner by the age of 22. On the outcomes side, the various components of labour market data were cleaned and recoded to generate measures of daily earnings, the proportion of the year spent in employment and the proportion of the year spent in receipt of active labour market benefits.

With more than 1.44 million observations, the analysis was undertaken for men and women separately – and a number of different model specifications were estimated. The results presented here illustrate the differential in labour market outcomes to each vocational qualification at Level 4 and below within the Regulated Qualifications Framework compared to the next highest level of qualification.

What did the analysis show?
The findings suggest that the association between vocational qualification achievement and labour market outcomes is generally positive and significant for most qualification aims at the different levels (for both males and females) – and particularly large for Level 4 vocational qualifications, Apprenticeships, and NVQs at Level 3 and 2.

Earnings differentials are particularly high for Level 4 vocational qualifications (37% and 26% for males and females respectively); Advanced Apprenticeships (40% and 21% compared to the level below and 25% and 16% compared to Intermediate Apprenticeships); NVQs at Level 3 (19% and 11%); Intermediate Apprenticeships (22% and 12%); but in contrast to a number of other studies, the analysis also illustrates a strong positive earnings differential for NVQ Level 2 qualifications (16% and 9%). The estimated earnings differentials for the other vocational qualifications at Level 3 and Level 2 are also typically positive (although smaller in magnitude), with the exception of males holding BTEC qualifications as their highest qualification (however, for many individuals BTECs act as a stepping stone for further study). Learners in possession of Level 1 vocational qualifications also attain positive earnings differentials, but the effect becomes smaller and insignificant once academic qualifications at Level 1 are controlled for (as the majority of these learners also hold 5 or more GCSEs at grades A*-G). The choice of the counterfactual seems to play a major role for lower level qualifications and further analysis in this area is currently under way.

Several previous analyses that have identified a relatively meagre earnings differential associated with intermediate and low-level vocational qualifications have nonetheless identified significant employment effects.[2] This analysis reinforces these existing studies – but interestingly finds that the association between employment and vocational qualifications is more important at lower levels of education. Specifically, the association between vocational qualifications and employment ranges between 5 and 10 percentage points at Levels 3 and 4 (relative to the baseline), and similar magnitudes are observed for L2 qualifications, while the association for L1 vocational qualifications is stronger.

For females, the employment differential ranges between 6 percentage points and 13 percentage points for Level 3 and Level 4 vocational qualifications, while it is mostly around 10 percentage points for Level 2 and Level 1 vocational qualifications.

The analysis also looked at the proportion of the year spent in receipt of active labour market benefits and the findings suggest that, males are less likely to claim benefits if they have higher levels of education. The magnitude is around 5 percentage points for higher-level qualifications and between 4 and 10 percentage points for lower-level vocational qualifications. For females the magnitude is around 10 percentage points for Level 3 vocational qualifications and between 7 and 15 percentage points for lower-level qualifications (all results should be interpreted as the reduction in the proportion of the year spent on benefits).

What’s next?

At this stage, there is limited comparability with traditional analyses using the Labour Force Survey, as the group of individuals covered by the matched LEO dataset is much younger on average and we report findings for daily earnings (as we don’t have information on hours worked and hourly wages). This means that people with higher qualifications might choose to work for longer hours and this would be captured in the estimates (although this employability effect would be partially taken into account as we look at daily rather than yearly earnings). A CVER research paper by Sheffield University and London Economics (forthcoming) looks in greater detail at the comparison across the two data sources.

Figure 1 Earnings differentials associated with vocational education and training

Source: London Economics’ analysis of LEO data

Key Stage 4 leavers in England in 2001/02 – 2003/04, with highest qualification achieved by 22. Outcomes at age 26 assessed in financial years 2012/13 to 2014/15. Dependent variable: Daily earnings. Controls for ethnic background, time elapsed since the learner left education, cohort dummies, FSM eligibility, SEN status, IDACI (income deprivation) score (all measured at Key Stage 4), and Key Stage 2 Maths and English test scores. Chart illustrates percentage daily earnings differential between treatment and counterfactual group (learners with highest qualification at the level immediately below). *includes Higher Apprenticeships, HNCs/HNDs and NVQs at L5.

[1] "The earnings differentials associated with vocational education and training using the Longitudinal Education Outcomes data" by Pietro Patrignani, Gavan Conlon and Sophie Hedges, CVER Research Paper 007 is available at
[2] See Dickerson 2005, London Economics 2011, and Buscha and Urwin 2013.

No comments:

Post a Comment