Adult learning and qualifications in Britain

Abstract

The importance of people gaining new, and high-level, qualifications in adulthood has been much emphasised in policy rhetoric. It is widely assumed that adults should engage in learning throughout their working life in order to adapt to changing conditions in the labour market and to ensure that national economies remain competitive in a global skills race. Educational researchers have frequently been rather sceptical about the numbers who actually achieve such upgrading in practice and have been critical of the feasibility and relevance of policies which attempt to address this issue. This paper provides empirical evidence on how many people acquire qualifications in adulthood, and whether they upgrade to higher levels of qualification than previously held, using British data from the 1958 National Child Development Study. Estimates are constructed of the volume of qualification acquisition and upgrading for this cohort through to age 50. On the basis of this new evidence, it is argued that previous analyses by educational researchers may have been overly pessimistic about the extent to which individuals engage in accredited learning over the adult lifecourse. The implications for research and policy are discussed.

Keywords:

• Qualifications
• upgrading
• adult learning
• lifecourse
• skills

Introduction

The importance of a highly skilled and well-qualified workforce to the modern economy has been repeatedly emphasised by policy-makers. The raising of citizens’ education levels is widely regarded as essential both for individuals to make progress in their own careers and for the health and vigour of the national economy in competitive global markets (OECD 2012). In the UK, there has long been a perception that the workforce is less well-qualified than many of its major competitors (Sanderson 1988; Wolf 1998). From the 1980s onwards international comparisons of the qualifications profile of the workforce consistently showed the UK in a poor light (National Commission on Education 1993; Layard, Robinson, and Steedman 1995). The next logical step was then to advocate the setting of targets for the improvement of qualification levels among the working-age population (Layard, Robinson, and Steedman 1995; Gorard, Rees, and Selwyn 2002) and this approach was taken up enthusiastically by policy-makers. Early in the present century, the Leitch Review of Skills (2006) was particularly important and influential. This recommended that the UK commit to becoming a world leader in skills by 2020, by benchmarking against the upper quartile of the OECD economies (Payne 2013). Targets were to be set across a range of levels in the hierarchy of qualifications. For example, the targets included exceeding 90% of adults qualified to at least Level 2, an increase from 69% in 2005. Progress towards the targets was monitored by a body called the UK Commission for Employment and Skills (UK CES). When the Conservative/Liberal coalition came to power in 2010, it quickly eschewed the use of formal target-setting. Nevertheless, since 2010 governments have continued to be committed to success in a ‘global skills race’ benchmarked against other OECD economies (Keep 2014). While policy levers, including levels and allocation of funding, may have changed substantially, the actual objectives of policy remained much the same.

Now, from an empirical perspective, a key question is whether individuals actually behave in the ways required for such aims to have any prospect of success in practice. That is, do people participate in the types of learning which will lead to qualifications, and are significant numbers of them likely to obtain qualifications at higher levels than they have previously held? While it is clear that more recent cohorts have remained in initial education for longer and acquired more and higher qualifications than previous generations, most of those who will comprise the workforce in, say, 10 or 15 years’ time are already in employment. There needs then to be a focus on so-called ‘second-chance education’ as well as initial education (Nordlund, Stehlik, and Strandh 2013). Significant progress in improving the qualifications profile of the workforce in the medium-term requires that adults must both obtain qualifications and upgrade to higher qualification levels. There is considerable disagreement on whether this has actually occurred. Some major studies by educational researchers have concluded, pessimistically, that few individuals gain qualifications in adulthood and that almost no-one upgrades to a higher level of qualification (Gorard and Rees 2002; Biesta et al. 2011). Some other studies have been somewhat more optimistic (Wilson and Bosworth 2006; Dorsett, Lui, and Weale 2011; Beblavy, Thum, and Potjagailo 2014). At present, then, the picture on the extent of qualifications upgrading by adults in Britain is far from clear.

This paper provides new evidence on the qualifications obtained by adults in Britain, and assesses the extent of upgrading over time based on analysis of a major longitudinal data source. By contrasting the pre- and post-2010 policy frameworks it also considers the effectiveness of policy in encouraging individuals to gain qualifications. In the next section, we review previous literature exploring the lack of consensus on the topic. Subsequent sections outline the data source, the construction of qualifications profiles for individuals and analysis of qualifications acquisition and upgrading. Implications for policy are discussed in the latter part of the paper.

Evidence review

Administrative data on course enrolments are plentiful (Ramsden 2006, 2010; HEFCE 2014) but do not tell us about whether qualifications were obtained, nor about whether individuals were obtaining higher level qualifications than previously held. For information on these topics, researchers have turned to survey data – either repeated cross-sectional or longitudinal surveys.

A group of social scientists at Cardiff University carried out several distinct but related projects on learning across the lifecourse in the late 1990s. The results of these projects were summarised in a large number of publications, notably in Gorard and Rees (2002) and Selwyn, Gorard, and Furlong (2006). It is convenient to refer to them here as the Cardiff Group. The broad picture which emerged from their work was of extended episodes of initial education amongst younger age groups. Young people were staying on in schools and colleges and gaining more, and higher levels of, qualifications on average than had been the case for earlier generations. But this was accompanied by an apparent decline in learning participation in later life. There was a risk that policy-makers were engaged in ‘robbing Peter to pay Paul’ by concentrating resources on learning in the early phases of the lifecourse (Gorard 2000; Gorard and Rees 2002). On the issue of qualification trajectories over the life cycle specifically, the analyses of the Cardiff Group were based on secondary analysis of cross-sectional data rather than results from their own surveys (Gorard, Rees, and Selwyn 2002). Using data for Wales from the Labour Force Survey for the years 1996 to 1999, they showed that insofar as there was progress towards larger numbers of the working population qualified to levels 2, 3 and 4 this was very largely, and perhaps wholly, due to well-qualified young people entering the workforce to replace older adults moving into retirement. They termed this ‘the conveyor belt effect’.

Many episodes of extended initial education are instead of, not additional to, the episodes of later-life learning that occurred for some in previous generations … there is no progress in participation and qualifications for adults while they are adults …. Each successive age cohort leaving initial education tends to have a longer education and a higher mean level of qualifications. However, each age cohort also tends to simply retain, rather than improve, their initial education over the remainder of their lives. Thus, the incidence of education among the population and the workforce as a whole increases, but only through the ‘conveyor belt effect’. (Gorard and Selwyn 2005, 1213)

More recently the Learning Lives project adopted a longitudinal approach to the analysis of learning over the lifecourse. It was funded by the ESRC’s Teaching and Learning Research Programme. This was a major project involving a large team of researchers from a number of British universities and utilising a range of methods, both qualitative and quantitative (Biesta et al. 2011). The quantitative component of the project used data from the British Household Panel Study (BHPS) to track participation in part-time, vocationally oriented training courses over a 14-year period from 1992 to 2005 amongst adults who were of working age at the start of the period (Macleod and Lambe 2008). The researchers looked at qualifications acquired during these part-time vocational training episodes, assigning levels to these qualifications and then comparing them to the highest level qualification that the individual previously held.

Over the 14-year period covered by their data about two-thirds of respondents did at least one vocational training course. Most participation spells did not lead to a formal qualification. There was some upward trend from about one-sixth of such spells leading to a qualification in the early 1990s, increasing to about one-quarter by 2004/2005. When participants did obtain accreditation as a result of a course, quite often the qualifications were not such as to fit into the National Qualifications Framework. That is they were in-house qualifications which were only relevant to the current employer and would presumably have no general value in the labour market. And when qualifications could be assigned to a point in the hierarchy, they were ‘almost invariably either at the same level or a lower level’ than the highest qualification already held by the respondent (Biesta et al. 2011, 35). A negligible proportion (less than 2%) of the qualifications obtained was at a higher level than those already held. These results were used by the Learning Lives Group to challenge the notion that adult students would progressively climb a qualifications ladder, with any new qualifications taken being at a higher level than earlier ones. It was argued that actual attainment in adulthood was very different such that trajectories of participation in formal episodes of learning would follow ‘complex and highly differentiated patterns’ (Biesta et al. 2011, 35–36).

The work of the Cardiff Group and the Learning Lives Group are the two major longitudinal studies of adult learning in Britain conducted by education researchers in recent years. Both reached the conclusion that qualification progression in adulthood had been very static with no, or almost no, upgrading. Moreover, both groups were very sceptical that policy could significantly boost the likelihood of more people obtaining qualifications. The Cardiff Group’s analysis was that people have already attained their desired qualification level in initial education and were therefore very unlikely to progress further in adulthood. In the analysis by the Learning Lives Group it would seem that people would be mainly interested in specific courses tailored to their own needs rather than generic upgrading of qualifications. It can therefore be inferred that there would be a very inelastic response to any change in adult learning policy which sought to increase the amount of upgrading.

Other researchers have been rather more optimistic about the extent to which adults increase their highest qualification level. For instance, the analysis of LFS by Wilson and Bosworth (2006) showed that there had been a large increase in the number of people in the labour market with formal qualifications. The predominant effect here had been a cohort effect due to young people who had qualifications at a higher level emerging into the labour force, but it had also ‘been reinforced to some extent by increasing qualification rates for older people as well’, which they refer to as an ‘upskilling’ effect.

Beblavy, Thum, and Potjagailo (2014) conducted an analysis of data for 27 European countries from the European Union Labour Force Survey (EU LFS). In the LFS data, it is not possible to follow individuals over time and so these researchers constructed a ‘synthetic panel data-set’ in order to consider the change in highest qualifications for various adult age groups in five-year steps to analyse changes over time within the same age groups. Specifically, they considered the proportions at medium and high levels of educational attainment as measured by the UNESCO International Standard Classification of Education. Hence, they identified movements from lower to upper secondary educational levels (referred to as ‘low education change’) and from upper secondary to tertiary educational levels (referred to as ‘high education change’). Their results show that, for instance, amongst the group aged 25–29 in 2000 about 5% in the UK had upgraded to a high level of educational attainment by 2010. This ranked the UK 13th from the 26 European countries with data. A similar proportion, roughly 5%, of the 30–34-year olds upgraded over the 10-year period between 2000 and 2010, with the UK ranked 9th among the 26 European countries here. Overall, then, the pattern of gaining higher level qualifications was found to be by no means static in the UK and quite favourable when compared to other European countries.

Researchers based at, or associated with, the LLAKES Centre at the UCL Institute of Education have conducted research on the benefits of gaining qualifications using BHPS data. The LLAKES team focused on adults, those aged from 25 to age 60, gaining qualifications between 1991 and 2007 (Dorsett, Lui, and Weale 2010, 2011; Evans, Schoon, and Weale 2013). Over a five-year period about 5% of men and 5% of women upgraded to a higher level of qualification than they had previously held. Over a longer period of time, sample size was considerably reduced by missing data but, among those for whom complete data were available, 14% of men and 17% of women upgraded to a higher level of qualification by 2007 than they had held in 1991.

To summarise the current state of the debate: the mainstream view among education researchers, based on the influential work of the Cardiff Group and the Learning Lives Group, has been highly pessimistic. In these accounts, rather few people obtained qualifications and miniscule proportions obtained them at a level which was above the highest level they had reached earlier on in their lives. However, other research appears to reach somewhat different conclusions, suggesting that the extent of upgrading by adults could, in fact, be markedly higher. This lack of consensus highlights certain limitations of the current evidence base. Firstly, it is very sparse with only a handful of empirical studies using British data. Secondly, some research has focused on rather short periods, often five years or less, as the time frame for analysis of upgrading. But it is apparent that the upgrading process takes a good deal of time and therefore it would be better if adults could be tracked for substantially longer periods of time. Thirdly, to date the evidence has been based on just two sources: the Labour Force Survey (LFS) and the British Household Panel Survey (BHPS). Now the LFS is not a genuinely longitudinal data-set and conclusions drawn from it may not be accurate, especially if there are changes in the composition of the sample over time. The BHPS is a longitudinal data-set but in order to follow people over time requires that respondents are present at each annual wave and this leads to enormous attrition from the sample. For example, while there are approximately 10,000 individuals in the survey, analysis over a 5-year period yields a sample of about 4000 in Dorsett, Lui, and Weale (2010, 2011) and only 432 men and 707 women in the longer term results of Evans, Schoon, and Weale (2013). In order to determine the extent to which adults are acquiring qualifications and improving their highest level of qualification, the best source of information is longitudinal data on individuals (Beblavy, Thum, and Potjagailo 2014). The sample needs to be large and representative. It needs to be capable of tracking over substantial periods of time, without grave loss of people from the sample, observing a sequence of learning episodes and whether they involve obtaining qualifications and movement up the qualifications ladder. In subsequent sections of this paper, new analyses are reported from a data source which meets these requirements.

Data and methods

Data preparation

Data were from the 1958 British birth cohort data-set, also known as the National Child Development Study (NCDS). This began as a perinatal mortality survey of every baby born in Britain in a single week in 1958. Originally, over 17,000 babies were in the sample. Follow-up data collection took place at several points in childhood up to age 16 and in adulthood at ages 23, 33, 42, 46 and 50. During the cohort members’ childhoods, data were collected by health visitors from the parents and from the children through educational and medical assessments. Some information was also gathered from teachers. In adulthood, data have been obtained directly from cohort members themselves via structured interviews. In 1978, exam results were gathered from all the schools attended by cohort members about the qualifications which they had obtained up to that point. This ‘Exams File’ provides objective information about qualifications obtained by the end of compulsory schooling at age 16 and in the four years immediately after that. Each wave of the NCDS in adulthood (from wave 4 at age 23 onwards) has asked respondents for quite detailed information about the qualifications they have acquired, usually since the previous wave. This information was used to map qualifications obtained, and the highest level of qualification, of respondents at ages 16, 23, 33, 42 and 50. The qualifications obtained by cohort members were coded to six levels, where each level is defined in terms of equivalency with National Vocational Qualifications (NVQs):

0 = no qualifications

1 = NVQ level 1 or equivalent, low-grade GCSEs or O levels

2 = qualifications at O level or GCSE A-C grade, NVQ level 2 or equivalent

3 = A level(s), NVQ level 3 or equivalent

4 = degree, NVQ level 4 or equivalent

5 = higher degree, NVQ level 5 or equivalent

Substantial changes have occurred to the education system in Britain during the lifetime of cohort members, including the replacement in England and Wales of O levels/CSEs by GCSEs and the introduction of NVQs and so assigning qualifications consistently within the hierarchy over time was not a simple matter. Previous work by Jenkins and Sabates (2007) and Dodgeon et al. (2011) was used to guide decisions about where qualifications should be placed in the hierarchy. It was useful to distinguish between academic (or general) and vocational qualifications and so qualifications profiles were disaggregated to construct a six-point scale for each of these two types of qualification. The qualifications obtained in compulsory schooling (O levels, CSEs and their Scottish equivalents) were assumed to be academic qualifications regardless of subject.

Although the NCDS data can be regarded as generally of very high quality, as with any longitudinal survey inconsistencies between different waves occur and it was necessary to resolve these. A basic assumption used here was that if someone was qualified to, say, Level 3 at age 33 they could not be at Level 2 later in life, that is individuals cannot lose their highest qualification. Yet, highest qualification did sometimes appear to move downwards over time. Cases where this occurred were first flagged as requiring further investigation. Then, in reconciliation work it was assumed that the Exams File, as an objective source, had priority in all matters where highest qualification levels were inconsistent. So, for example, respondents who were at Level 2, say, in the Exams File but had Level 0 or 1 given as their highest ever qualification in waves 4 or 5 were recoded so that their highest level qualification was Level 2. There were also a few cases where individuals appeared to have moved from low-level qualifications at one wave to notably higher qualifications at the next wave. These were checked for plausibility on a case by case basis. This involved checking against the employment histories available in NCDS to see whether the respondent had spent substantial amounts of time in full- or part-time education in the period immediately before they gained new, higher level qualifications. For the majority of such cases it did appear to be plausible that they had obtained the qualifications, but a handful of improbable cases were recoded so as not to have obtained the high-level qualifications claimed.

The objective of the research was to track the qualifications obtained up to the age of 50. So the main constraint, in terms of sample size, was that people must have been present in wave 8, the age 50 wave of the survey. People who dropped out permanently before then were, of necessity, excluded. There were 9790 cases present at the age of 50, which imposed the upper limit on sample size. In order to construct a full qualifications profile from age 16 through to age 50 the ideal situation would be that respondents were present in the 1978 Exams File and in all 5 adult waves of the survey. There were some 5524 cases meeting that criterion. However, for many people who were absent at one or more waves of the survey, but had returned by age 50, it was still possible to derive a full qualifications profile. This was done using later waves to address data missing at one or more previous waves. For example, in wave 5 (age 33) respondents were asked about qualifications attained before and after age 23. Hence for those respondents missing at wave 4 (age 23) but present at wave 5, patterns of attainment and qualifications achieved by age 23 could be recorded using the wave 5 data. This approach was applied to other waves too. By this process, the number of cases for which a full qualifications profile could be constructed was increased to 8939 and this was the sample used to derive all the results shown in this paper.

Defining terms

The phase of adult education begins once initial education has been completed. There is scope for debate about when exactly initial education comes to an end. It is usually assumed to be somewhere in the early or mid-1920s. Since NCDS cohort members were interviewed at age 23, it is convenient to take that age as the terminal point of initial education. The phase of adult education therefore occurs from age 23 onwards and, with the data available in NCDS, it is possible to observe whether people obtain qualifications between the ages of 23 and 50. This period of adult education can be broken down into an immediate post-initial phase, from ages 23 to 33, and mid-life learning which occurs between ages 33 and 50. The definitions to be used are:

Age 16 – End of compulsory schooling

All education up to age 23 – Initial education

All education from age 23 onwards – Adult education

Ages 23 to 32 inclusive – Post-initial education

Ages 33 to 50 – Mid-life learning

Results

Obtaining qualifications in adulthood

During the adult education phase of the lifecourse (ages 23–50) some 71% of the sample obtained at least one qualification, and 52.5% did so in mid-life (ages 33–50). The proportions obtaining one or more qualifications in each phase of the lifecourse after compulsory schooling through to age 50 are shown in the last column of Table 1.

Table 1. Obtaining qualifications during different phases of the lifecourse.

	Academic qualification (%)	Vocational qualification (%)	Any qualification (%)
Aged 17–22 incl.	37.6	36.8	61.0
Aged 23–32 incl.	16.1	36.1	45.0
Aged 33–41 incl.	12.9	30.3	36.0
Aged 42–50 incl.	5.3	29.5	32.3
N: 8939

It can be seen that, as the cohort grew older, there was a gradual decline in the proportions obtaining qualifications. The decline was much more marked for academic qualifications – from about 16% who obtained a qualification between ages 23 and 32 compared to only around 5% between ages 42 and 50 – than for vocational qualifications. Almost 30% of cohort members obtained a vocational qualification between ages 42 and 50. The picture that emerges is, nonetheless, of substantial numbers having obtained qualifications in the adult phases of the lifecourse, with a strong tendency for these qualifications to be mainly vocational, rather than academic.

The breakdown by gender in Figure 1 reveals that, amongst this cohort, men were more likely than women to obtain qualifications in the early phases of adulthood (up to age 33) but that women were more likely than men to obtain qualifications in mid-life, between the ages of 33 and 50. Between the ages of 17 and 22 over two-thirds of men obtained a qualification while about 55% of women did so. In mid-life 48% of men and 57% of women obtained at least one qualification.

Figure 1. Obtaining qualifications across the adult lifecourse to age 50, by gender.

Upgrading

At age 16, the age at which compulsory schooling came to an end for this cohort, nearly a quarter (23.2%) had no qualifications at all. As can be seen in Figure 2, this proportion had fallen to less than 10% by age 33, and to just under 6% by the age of 50.

Figure 2. Proportion with no qualifications in NCDS to age 50.

But it was not just in the reduction of people with no qualifications that progress was made in adulthood. There was steady upward progress, across all the levels of the qualifications hierarchy, as the cohort aged (Table 2). The percentage of people whose highest qualification was at Level 2 or below fell from 62% at age 23 to 55% at age 33, and to 46% by age 50. Conversely, those with qualifications at Level 4 and above comprised 21% at the age of 23; nearly 29% by age 33, and close to 37% by the time the cohort members were 50 years old.

Table 2. Highest overall level of qualification, by age.

	Age 23 (%)	Age 33 (%)	Age 50 (%)
No qualifications	13.4	8.7	5.9
Level 1	14.7	13.9	11.5
Level 2	34.3	32.5	28.8
Level 3	16.7	16.3	17.1
Level 4	20.2	17.8	21.5
Level 5	0.7	10.8	15.2
ALL	100.0	100.0	100.0
N: 8939

Discussion

Educational trajectories over the lifecourse constitute an important topic for social science and policy researchers, with implications for individual progression during careers as well as inequality by social background. Previous literature has tended to concentrate largely on young people making initial transitions into further and higher education (Pallas 2004; Fuller 2007). In Britain, there have been only a handful of studies analysing participation and non-participation in learning during the adult lifecourse.

This paper contributes to the longitudinal analysis of education in adulthood. Focusing on qualification trajectories it has been argued that previous longitudinal studies may have been too pessimistic about the extent to which people acquire qualifications in adulthood. Data from the 1958 cohort suggest that, amongst a sample of almost 9000 respondents over half obtained a new qualification in mid-life, between the ages of 33 and 50. Moreover, some 26% of the sample raised their highest level of qualification between the ages of 23 and 32, while 18% did so between the ages of 33 and 50. In whichever way the data are summarised then, the picture is a much more dynamic one than might have been expected given the very static results which have emerged from some previous analyses such as Gorard and Rees (2002) and Biesta et al. (2011).

The main theme of the results presented here is that many people in the 1958 birth cohort returned to study after the completion of initial education, undertaking courses which led to qualifications. Why did they do so? Most of the qualifications were vocational, suggesting that career development was a key motivation for study. There have been major changes in the types of skills in high demand since this cohort left initial education (Smith 2009). The years since then have seen a huge growth in service sector employment. Long-term employment within the same organisation has become increasingly uncommon and workers have recognised the need to develop new skills throughout their careers. In Britain the proportion of jobs requiring graduate qualifications rose from 10% in 1986 to 26% by 2012, while jobs requiring no qualifications declined from 38% to just 23% over the same period (Felstead et al. 2013). Changes in the demand for skills since they left school might well have motivated members of the cohort to seek additional qualifications.

Gender differences in patterns of obtaining qualifications over the lifecourse were striking. In initial education and in early adulthood (up to age 33) men were more likely to obtain qualifications than women, but in mid-life markedly higher proportions of women obtained qualifications than men. In mid-life women were also much more likely to upgrade to higher levels of qualification than men. Amongst this cohort women had tended to fall behind men, on average, particularly after the completion of compulsory schooling. The nature of the youth labour market in the mid-1970s and prevailing assumptions about gender roles meant that women were less likely than men to have acquired intermediate vocational qualifications in their 20s and early 30s. But there was clearly a strong tendency for women to catch up on educational experience and qualifications in mid-life.

Because it can take many years for people to gain new qualifications, it is important to track individuals over long periods of time and this can be accomplished using cohort data. In this study, we have looked at qualifications in adulthood for more than a quarter of a century, from when cohort members were aged 23 to when they were 50 years old. This long-term perspective may help to explain why the results here contrast with some previous research which has tended to study people over shorter periods of time. Some earlier studies have also focused on specific qualifications, for example degrees, rather than qualifications in general. Or rather narrow definitions of learning have been adopted such as looking first at whether people were doing particular training courses and then seeing whether they get some qualification at the end of their course (Biesta et al. 2011).

The two major longitudinal studies of adult learning in Britain conducted by education researchers in recent years are the work of the Cardiff Group and the Learning Lives Group. Both reached the conclusion that qualification progression in adulthood had been very static with no, or almost no, upgrading. The implication was that policy was largely irrelevant. It was argued, for instance, that the lack of people acquiring qualifications and moving to higher levels in the qualifications framework as adults demonstrated that ‘front-loading’ into initial education weakened opportunities and incentives to return to education later on in life and that policy initiatives were unlikely to change this. In the analysis by the Learning Lives Group, it would seem that people would be mainly interested in specific courses tailored to their own needs rather than generic upgrading of qualifications. It can therefore be inferred that there would be a very inelastic response to any change in adult learning policy which sought to increase the amount of upgrading. However, the findings in this paper cast considerable doubt on the notion that adults simply do not acquire qualifications or upgrade to higher qualification levels. For much of the period through to 2008 when this cohort reached the age of 50 it can be argued that both labour market conditions and the policy environment were reasonably favourable for adults wishing to gain new qualifications (Fuller 2007). Against that background adults were able to gain qualifications in considerable numbers.

But, in recent years, the economic situation and policy towards adult learners have become much harsher. There has been a drastic decline in the number of part-time mature students in higher education. Between 2010/2011 and 2013/2014 the proportion of UK/other EU undergraduate entrants to English higher education institutions and further education colleges studying part-time fell from 40 to 27% (HEFCE 2014). A good deal of this dramatic decline can be attributed to reforms to the funding of higher education in England, which replaced grant funding with financing through loans and effectively trebled the cost of HE study. Extension of loans to some part-time students was intended as a way of levelling the playing field between full-time and part-time study. However, in practice, increased fees coupled with debt aversion among older adults led to very substantial reductions in the number of part-time and mature students in HE (Callender 2014). The decline was exacerbated by the recessionary economic climate of unemployment, low pay and a squeeze on employer training budgets. Other types of learning have also been hard hit by the recession and by cuts to funding. The total number of adult learners participating in government-funded further education fell from 3.5 million in 2009/2010 to 2.9 million by 2013/2014 (Skills Funding Agency 2015). Over this period there was a 22% drop at Level 2, GCSE or equivalent, and a 19% decrease at Level 3, A-level or equivalent. Although it is still too early to be certain of the consequences for the extent of long-term upgrading of qualification levels amongst adults these sharp downturns in participation must be of grave concern.

Conclusion

Data from the 1958 cohort suggest that, amongst a large sample of almost 9000 respondents, some 71% obtained a qualification between the ages of 23 and 50, and that 52.5% did so in mid- life, between the ages of 33 and 50. Moreover, there was considerable evidence of progression up the ladder of qualifications. Movement between levels is often summarised by the crossing of particular thresholds which are regarded as important in policy terms. Some 2509 cases from the sample were below Level 2 at age 23. By age 33, almost 20% of these had reached at least Level 2 and by age 50, 38% had done so. Higher up the spectrum of qualifications, among people who were at Level 4 or above at age 50, 57% had already attained this level by age 23, but some 43% had reached this level only in the post-initial (ages 23 to 33) or mid-life (ages 33 to 50) phases of the lifecourse.

The prevailing view, the dominant narrative, amongst educational researchers has been that few individuals gain qualifications in adulthood and that almost no-one upgrades to a higher level of qualification (Gorard and Rees 2002; Biesta et al. 2011). In whichever way the data are summarised, the picture here is a much more dynamic one than might have been expected given the very static results which have emerged from previous research.

Now, the 1958 cohort obtained their qualifications within a policy environment which may be characterised as broadly favourable to adult education. Since 2010 the policy climate has become a good deal frostier. As a consequence, participation in further education and the numbers of part-time mature students in higher education have fallen dramatically. It takes a long time for patterns of upgrading to become fully apparent, but it cannot be assumed that younger cohorts will continue to engage in learning and obtain qualifications to the same extent as they move through their 30s and 40s. If the upskilling of significant proportions of the workforce remains an objective, current policies are very poorly aligned for the achievement of this goal.

Notes on contributor

Andrew Jenkins is a senior research officer in the Department of Social Science, UCL Institute of Education and a visiting research fellow at the Policy Institute, King’s College London. He specialises in the application of quantitative methods in educational research, and especially the analysis of large-scale longitudinal data-sets. His substantive research interests are mainly about learning in adulthood, including participation in different types of learning during distinct phases of the life course, the effects of learning on labour market outcomes, and the mental health and well-being benefits of learning for older adults.

Disclosure statement

No potential conflict of interest was reported by the author.

Funding

This research was supported by a British Academy mid-career research fellowship.

Acknowledgements

I am most grateful to the British Academy for their generous support of my research. I owe a very large debt of gratitude to Professor Dick Wiggins who acted as advisor throughout the project. I wish also to acknowledge the creators of the NCDS data resource, the Centre for Longitudinal Studies, UCL Institute of Education and the UK Data Service for providing access to data.