Steve McIntosh, from the Department of Economics at the University of Sheffield, is an applied labour economist, with a particular focus on education issues and how they relate to labour market outcomes. Here he asks why the focus on apprenticeships in the recent election, and what are their apparent benefits?
The basic methodology shared by all research in this area looks at the wages and employment rates of those who complete an apprenticeship, compared to the same outcomes for a comparison group, controlling for as many other wage-deterrmining characteristics of the two groups as possible. Where different studies vary is in their choice of comparison group. Those studies that use survey data, such as the Labour Force Survey, typically compare apprentices to those who have not undertaken an apprenticeship, and whose highest qualification is one level below. For example, for advanced (level 3) apprentices, the outcomes of those with a level 2 qualification as their highest is taken as the estimate of what would have happened to the apprentices if they had not undertaken their apprenticeship. Those studies that use administrative data often compare the outcomes of those who complete an apprenticeship to those who undertake an apprenticeship but do not complete.
The headline figures for the estimated returns are encouraging, and justify the attention given to apprenticeships by the political parties. Before I present the results, however, it is important to recognise that they could be over-estimates of the true effect of apprenticeship training. Because many apprenticeship programmes are oversubscribed, employers can choose from the list of applicants, and are likely to select the best candidates in terms of ability and motivation. The individuals we observe undertaking an apprenticeship might therefore be the sort of people who would have done well in the labour market anyway. Similarly, when comparing those who complete an apprenticeship to those who begin an apprenticeship but do not complete, the completers are likely to be of higher ability, and so again would likely have earned more even without the apprenticeship. In both cases, some of the observed wage differential could therefore be due to the characteristics of the apprentices, rather than the apprenticeship training itself. Future CVER work will investigate ways to control for any such biases. For now, though, we can say that the estimated returns, below, are probably large enough to accommodate partly being due to favourable learner characteristics, and still yield a worthwhile return to training.
the results say
The earlier studies were more likely to be ones that use survey data. For example, McIntosh (2007) used the Labour Force Survey for 2004-5, and found individuals whose highest qualification was an apprenticeship at Level 3 earned on average 18% more than those who highest qualification was at Level 2, while for a Level 2 apprenticeship, the wage premium was 16% compared to individuals whose highest qualification was at Level 1 or 2. Summing over the working lifetime, taking into account the probability of being in employment at each age, and also the costs of apprenticeship training, the net value of obtaining an apprenticeship in current monetary terms was found to be £105,000 for Level 3 apprenticeships and £73,000 for Level 2 apprenticeships.
This research was updated by London Economics (2011), using Labour Force Survey data for 2004-9, with very similar results. The observed wage premia were 22% for a Level 3 apprenticeship and 12% for an apprenticeship at Level 2. This produced net present value results for the lifetime benefits of between £77,000 and £117,000 for Level 3 Apprenticeships and between £48,000 and £74,000 for Level 2.
More recent estimates of the labour market benefits of apprenticeships have used administrative data, matching data from the Individualised Learner Record (ILR) to wage information from HRMC. The results obtained are reassuringly similar to those produced when using survey data such as the LFS. London Economics (2013) report one such analysis, estimating wage premia in the years following completion of an apprenticeship. They find that an individual who obtains a Level 3 apprenticeship earns 15-20% in each of the first six years following the completion of the qualification, relative to an individual who fails to complete. At Level 2, the wage premium to an apprenticeship is around 10-15% in each year following completion, except for the first year when the initial return is much higher. A similar analysis of apprenticeship completers versus non-completers at the same level was undertaken by Bibby et al. (2014), using a similar matched data set covering the period 2002-2012. They report a 16% wage premium for Level 3 apprenticeship completers and an 11% wage premium for Level 2 apprenticeship completers, averaged over the period 3-5 years following the completion of the apprenticeship.
Variation in returns
Studies therefore suggest large individual wage benefits to undertaking and completing an apprenticeship, irrespective of which data set, methodology and time period are studied. It is such results that have been used to justify the extension of the apprenticeship programme. However, it must be noted that these results do not mean that all apprenticeships have high payoffs to the individuals and to the exchequer. We must recognise that these figures are averages, and that there will be variation around these averages. Further analysis has investigated the characteristics of apprenticeships that are associated with high and low returns. The framework area of an apprenticeship has been shown by a number of the studies mentioned above to be strongly associated with the estimated wage premium. Construction and engineering apprenticeships have consistently been shown to have higher wage premia, though their relative strength is slightly lower in a full cost-benefit analysis, once the higher cost of provision of such apprenticeships is also taken into account. At the other end, apprenticeships in the areas of retail, hospitality and tourism are often observed to yield small and insignificant returns. In addition, though not studied specifically with respect to apprenticeships, previous studies such as Bibby et al. (2014) show higher returns to vocational qualifications at all levels when they consider 18-24 year olds than for 25+ year olds. It is therefore of some concern that the big increase in apprenticeship numbers since 2010 is for the 25+ age group, with a smaller increase for the 19-24 age group, and no increase at all for the 16-18 year old age group (see Figure 1). Similarly, data by sector subject area show that the areas seeing the largest increases in apprenticeship numbers over the previous decade have been retail and caring, together with business administration. Engineering apprenticeship numbers have seen a much smaller increase, while numbers in construction are falling. We therefore cannot assume that the average wage returns observed in earlier periods, as described above, will be earned by many of the additional apprenticeships that have been observed in the recent growth period.
Therefore, while the benefits to apprenticeships are, on average, very pleasing, reflecting the value of the training received, and the growth in apprenticeship numbers is also a good sign, it is important that apprenticeship quality is maintained or even enhanced, that attention should be prioritised where value is greatest, and that in the efforts to increase apprenticeship numbers, the brand is not diluted by classifying ongoing training, specifically of older workers who are already employees, as apprenticeships. Apprenticeships must remain an intensive training experience, of significant duration, and providing the apprentice with a new set of skills with which to practise in their chosen profession, and not become just a validation of existing skills.
Figure 1: Apprenticeship numbers over time by age
Source: SFA, Statistical First Release SFA/SFR29