Returning from earning: UK graduates returning to postgraduate study, with particular respect to STEM subjects, gender and ethnicity
Steve d’Aguiar (formerly of Higher Education Funding Council for England)
Neil Harrison (Faculty of Arts, Creative Industries and Education, University of the West of England)
Abstract: It has been argued by some (e.g. the Confederation of British Industry [CBI]) that graduates lack the skills that render them employable. In particular, graduates of science, technology, engineering and mathematics (STEM) subjects are often portrayed as being unready for the world of work.
This paper uses three large-scale national datasets from the UK to explore this assertion, including the results of the Destinations of Leavers from Higher Education surveys. It reports analysis of 22,207 individuals who graduated from their first degree in 2007, and works from the hypothesis that those entering the workforce and then returning for taught postgraduate study are primarily doing so due to underemployment in the period following graduation.
The study uses binary logistic regression and finds that a range of educational, demographic and employment-based variables have a significant relationship with the propensity to return for taught postgraduate study. Of particular note, those returning tend to be high-achievers from elite universities in low-skill work after graduation, as well as women and those from minority ethnic communities; this suggests a mix of individual and structural factors at work. In addition, STEM graduates were significantly less likely to return, apparently challenging the argument advanced by the CBI.
Keywords: higher education, STEM, work-readiness, underemployment, DLHE
Background
An ongoing discourse around a perceived under-preparation of graduates for employment continues, both in the UK (e.g. Moreau and Leathwood 2006; Tomlinson 2007; Holmes 2013; Tymon 2013) and other developed countries (e.g. Martin et al. 2000). This discourse generally makes reference to a set of so-called ‘soft’ skills, such as the ability to present ideas and interact with colleagues (Leitch 2006; The Royal Society 2006; Henderson et al. 2010). Concerns have been expressed, notably from the Confederation of British Industry (CBI), that this under-preparation is particularly true of graduates in science, technology, engineering and mathematics (STEM) subjects (CBI 2008; Smith 2007).
However, this appears to be challenged by the UK Commission for Employment and Skills (UKCES) and their Employer Skills Survey conducted in 2011 (UKCES 2012a,b). This involved over 87,000 telephone interviews with diverse businesses across the UK, and investigated, inter alia, graduates’ readiness for work and skill shortages. Only 14 per cent of English employers considered that those leaving higher education were poorly prepared for work (UKCES 2012a,b). Similarly, the Institute of Directors reported that just 9 per cent of respondents to a survey of 500 directors were dissatisfied with graduates knowledge and skills (IoD 2007). Indeed, evidence for the under-preparedness of graduates appears to be largely anecdotal (The Royal Society 2006; Smith 2007; Henderson et al. 2010), although employers continue to feel that universities could do more (Lowden et al. 2011).
The particular interest in STEM subjects derives from a widely-held belief that they are ‘vital for our economy to enable the UK to do well as a nation’ (House of Lords 2011, 1); similar sentiments are also expressed by Smith (2009). Both Henderson et al. (2006) and Smith (2007) report that employers believe that modularised learning in universities has contributed to a decline in graduates’ scientific knowledge, such that the quality of STEM graduates has fallen, rather than the quantity. Mellors-Bourne et al. (2011) report a perception that many STEM graduates are lacking in core scientific knowledge and workplace experience, with the CBI’s survey of senior executives revealing that ‘42% of firms consider STEM graduates lack the right skills’ (2008, 7).
There are also concerns that there are not enough graduates in STEM subjects and, in particular, groups such as women (Roberts 2002; Kirkup et al. 2010; Regan and Dillon 2012; Smith 2012), some ethnic minorities (Elias et al. 2006; DeWitt et al. 2011; Smith and White 2011) and those from lower socioeconomic backgrounds (Smith and White 2011; Campaign for Science and Engineering 2012) are under-represented on many (especially high status) courses, although women now outnumber men overall in STEM subjects (Ratcliffe 2013).
Given the policy interest around the preparation for employment, relatively little empirical work has been undertaken. The focus here is, therefore, on the analysis of secondary national quantitative data, with particular reference to the hypothesis that STEM graduates are, on average, less prepared for graduate employment than their peers who studied non-STEM subjects. It uses data from the Destinations of Leavers from Higher Education (DLHE) surveys, combined with student-level data held by the Higher Education Statistics Agency; there has been little previous research conducted using the DLHE dataset, which is collected from graduates six months and three-and-a-half years after completion, notable exceptions being Chevalier (2008), Gittoes (2009), Norton (2008) and Woodfield (2011).
Problematising work-readiness and underemployment
This study is situated within the broader discourse around graduate employability. Space precludes a full discussion of this much-debated concept (see Holmes 2013 for an historical overview), but a commonly-used definition is ‘a set of achievements – skills, understandings and personal attributes – that makes graduates more likely to gain employment and be successful in their chosen occupations’ (Yorke 2004, 8). Here we primarily engage with a more limited (and recent) concept of ‘work-readiness’, in the sense of students being able to ‘hit the ground running’ without the need for additional training in the initial period of employment (Mason, Williams and Cranmer 2009). As evidenced by the CBI and UKCES reports cited above, employers argue for the value of work-readiness in reducing costs, maintaining continuity and increasing productivity (Boden and Nedeva 2010). Discussions have mainly focused on the ‘soft’ skills associated with the contemporary workplace, particularly around working with other people and marshalling and communicating information and ideas. It is also asserted that there are discipline-specific elements of work-readiness, for example in healthcare (Walker et al. 2013), accountancy (Jacklin and De Longe 2009), engineering (Jollands, Jolly and Molyneaux 2012) and science (Coll and Zeegwaard 2006). Specifically, in relation to STEM graduates, there are concerns about cross-cutting science skills around scientific principles, laboratory practice and applied mathematics (Roberts 2002; Henderson et al. 2010).
Within this concept of work-readiness, graduates are constructed as needing both academic credentials and a wider (and, arguably, ill-defined – see Hinchliffe and Jolly 2011 or Tymon 2013) set of skills, knowledge, behaviours and dispositions that simultaneously contribute to their ‘attractiveness’ to employers. Employers therefore expect graduates to be able to evidence their work-readiness through the information provided in job applications, through the interview process (and associated test exercises) and through their performance once in post. A weakness would therefore hamper the individual’s ability to secure work or to progress at the expected pace in the early stage of their career. In addition, graduates may possess different levels of career management skills and therefore have suboptimal approaches to marketing themselves to employers (Bridgstock 2009).
A lack of such work-readiness could occur across whole programmes, subject areas or universities (the publication of comparative statistics being partly the rationale behind the DLHE), but also at the individual level, with graduates achieving a strong academic result, but failing to acquire the skills or specialist knowledge valued by employers; Mellors-Bourne et al. (2011) provide some evidence for this. In other words, a graduate’s work-readiness can be enhanced or compromised by features of their degree programme (e.g. insufficient content focused on contemporary employment or opportunities to develop practical skills within the curriculum) or by their own interaction with it in terms of the skills and knowledge resulting (Cranmer 2006; Mason, Williams and Cranmer 2009). Rather than being a binary condition, work-readiness can more usefully be seen as a form of positional capital, but, in the sense used here, excludes other potential positional advantages such as the prestige attached to the university attended.
Work-readiness may therefore feed into (but not define) the ability of graduates to secure work at an appropriate level and thrive within this environment in the first few years of their career; this is a phenomenon that is widely-recognised by students themselves (Moreau and Leathwood, 2006; Tomlinson 2007, 2008; Tymon 2013). In a highly-competitive labour market, low work-readiness could therefore compromise the potential to secure or maintain graduate-level employment, leading to the individual becoming ‘underemployed’ in a role that is less skilled than their degree (or degree class) would suggest.
Indeed, evidence for widespread underemployment is strong. Chevalier and Lindley (2009) compared two cohorts of students spanning a period of rapid expansion in higher education in the UK: those who graduated in 1990 and in 1995. They concluded that the proportion working in roles not requiring a degree nearly doubled between the two cohorts (from 7 percent to 11 percent) and that there was a wider group of graduates in jobs only recently requiring degrees. Drawing on their longitudinal study of 17,000 undergraduates, Purcell et al. (2012) estimate that as many as 40 percent spent significant periods in what they define as ‘non-graduate work’ in the two years after university, while Mosca and Wright (2011a) provide a figure of 37 percent six months after graduation, for the 2002/03 cohort. Smetherham (2006) provides similar evidence based on the perceptions of graduates, finding around one-in-five felt that they were qualified for, and capable of, more demanding work. Indeed, Scurry and Blenkinsopp’s (2011) review of the literature suggests that there are both objective and subjective elements to underemployment – a sense in which some jobs are inherently ‘non-graduate’ (see Elias and Purcell 2004), while other jobs may cause the graduate to feel that they are working below their qualification or ability level.
Poor work-readiness could therefore be one explanation for underemployment. However, this argument has its critics. Moreau and Leathwood (2006) argue that an over-focus on the individual’s work-readiness is a misframing, as it fails to respect the structural inequalities that provide differential access to graduate jobs. This is supported by Wilton (2011), who uses a sample of business graduates to demonstrate a disjuncture between self-reported employability skills and the likelihood of underemployment after graduation. Others (e.g. Boden and Nedeva 2010) point out that complaints about graduate skills has been a popular refrain from employer groups for several decades and suggest that this discourse may reflect more about power relations between employers, government and universities than the abilities of individual students. There are also concerns about the subjective and unmeasurable nature of employers’ requests (Hinchliffe and Jolly 2011, Tymon 2013).
More broadly, Brown (2003) argues that after a period of post-industrial realignment in the UK (and elsewhere in the developed world), the number of highly-skilled jobs has stagnated, producing an ‘opportunity trap’ in which an ever-growing pool of graduates emerging from an expanded higher education sector is funnelled into a labour market of fixed size. With the supply of graduates significantly outstripping the demand from traditional employers, an highly competitive labour market is established, especially for the most prestigious and lucrative jobs. In this environment, even the possession of a first class degree does not guarantee employment at the desired level, where employers are able to filter candidates ruthlessly by various forms of positional advantage, including work-readiness. The net result is a large surplus of qualified graduates finding themselves displaced into non-graduate work and ‘institutionally disappointed’ (Brown 2013); this may occur within a particular field (e.g. forensic science, which produces more graduates each year than the total jobs in the discipline: Tobin 2011) or geographical location (see Mosca and Wright 2011b). In the long run, this ‘opportunity trap’ sees previously non-graduate occupations redefined as graduates come to predominate and blurring the lines between the objective and subjective forms of underemployment (Scurry and Blenkinsopp 2011).
In the short-term, it is important to consider which groups are more likely to find themselves positionally disadvantaged. Wilton (2011) finds that women and graduates from minority ethnic communities report the greatest skills gains from higher education, but the poorest early career experiences in terms of level of work, satisfaction and pay. In similar vein, Moreau and Leathwood (2006) argue that factors such as gender, ethnicity, age and disability status may trigger active (e.g. preferential employment of certain groups) and passive (e.g. an unconscious drive to recruit more ‘people like us’ who will ‘fit in’) forms of discrimination, regardless of their work-readiness. Furthermore, Cranmer (2006) finds that reputational status is held to confer significant advantage to those passing through elite universities; a position supported by Boden and Nedeva (2010), who argue that lower status universities strive to substitute this disadvantage with work-readiness initiatives. This reputational factor is likely to further disadvantage graduates from minority ethnic communities who continue to be under-represented in elite universities. These critical voices therefore argue that structural inequalities have more role in defining patterns of graduate employment than work-readiness.
Holmes (2013) rejects both the work-readiness approach and the critical response as simplistic and poorly grounded in empirical data. Instead, he proposes a ‘processual’ model which stresses a negotiated transaction between graduate and prospective employers, in which the former attempts to convince the latter that they fit their definition of ‘graduateness’. This will inevitably differ between employers and may be informed by the structural inequalities discussed above, but it also encompasses aspects of the graduate themselves, what they are able to evidence and how they present themselves. This finds support in Tomlinson’s (2007) distinction between ‘careerists’ and ‘ritualists’, whereby the former seek to accrue positional advantage (e.g. through experiences outside of the formal curriculum) in contrast to the latter who are unwilling to ‘play the game’ and resign themselves as a result to lower status employment, at least in the first instance. Furthermore, Tymon (2013) notes that some of the ‘skills’ identified as supporting employability are actually personality constructs and may therefore be immune from the very concept of skills development.
Under Holmes’s (2013) ‘processual’ conceptualisation, success in the graduate labour market might reasonably be seen as function of both work-readiness and the ability and willingness of the individual to compete in the marketplace, partly through their qualifications, but also through how they engage with a range of structural and agentic challenges. As a result, there is a danger that graduates may (a) fail to secure work in their chosen career, (b) find themselves in work at a level that they feel is too low, and/or (c) fail to progress within their career at an appropriate rate. Indeed, Holmes (2013) predicts a group of such individuals which he defines as having a ‘failed identity’ in the graduate labour market. We now consider what responses they might have to this situation.