Jump to ContentJump to Main Navigation
The success paradoxWhy we need a holistic theory of social mobility$

Graeme Atherton

Print publication date: 2016

Print ISBN-13: 9781447316336

Published to University Press Scholarship Online: September 2016

DOI: 10.1332/policypress/9781447316336.001.0001

Show Summary Details
Page of
Subscriber: null; date: 26 February 2017

The shape of the labour market: hourglass, diamond or molecule?

The shape of the labour market: hourglass, diamond or molecule?

(p.109) Six The shape of the labour market: hourglass, diamond or molecule?
The success paradox

Graeme Atherton

Policy Press

Abstract and Keywords

This chapter looks at the relationship between the economic changes affecting contemporary labour markets and the idea of holistic social mobility being developed through this book. It explores the rise of the hourglass economy. It argues though that the hourglass concept however, may however be actually a rather simplistic view of the economy useful to capture some broad changes but less useful at really describing the complexity of the changes that are underway. The chapter considers how looking at the labour market more forensically and in the context of what progression means both within and between jobs in the broader sense the labour market may resemble a pyramid or a molecule. In this more three dimensional conception of the labour market it is less easy to say what sort of mobility i.e. downward, upward or sideways, movement between jobs represents and how different jobs relate to each other.

Keywords:   hourglass pyramid, molecule

The current changes in the labour market could have potentially seismic impacts across the whole field of social mobility. They are having an impact on both the type of jobs available and their nature in terms of the tasks associated with particular roles, including the knowledge-intensive jobs. These changes fatally undermine the idea that politicians can engineer social mobility by raising the attainment of children and hence educate their way to a more equal society.

The combination of technology and ideology driving these changes is also changing the very nature of the occupational stratification systems on which the academic study of social mobility depends. The relationship between different types of job is what underpins occupational stratification systems. There need to be clear dividing lines between jobs. When these lines become blurred and when within job categories there becomes increasing differentiation in levels of skill required, tasks performed and income levels these stratification systems become more less meaningful as indicators or social division.

In this chapter, the rise of the ‘hourglass economy’ is explored – as the number of mid-range jobs disappears, this leaves us not with a linear form of occupational structure, but with something more akin to an hourglass shape. This concept, however, is a rather simplistic view of the economy – useful to capture some broad changes, but less useful at describing the complexity of the changes that are currently underway. If an accurate picture of the 21st-century labour market is to be portrayed, we may need more mixed metaphors.

These changes in the labour market strengthen the case for a holistic perspective on social mobility, but they also demand an understanding of the impact they are having. It is not just the type of jobs that are changing; it is also the very meaning of work that is under question. Success and progress via work are the bedrock of the present social mobility discourse, and these fundamentals must themselves be examined. Have we gone too far in the fetishisation of work? Is it desirable – or even possible – to step back to consider how we re-position work in relation to leisure and family? And if we do, what will that mean for social mobility?

(p.110) The ‘hourglass economy’

There is a welter of evidence to support the argument that the seeds of the changes in the economy and the labour market now being experienced were sown in the late 1970s and early 1980s (Barnet and Cavanagh 1994, Michie and Grieve Smith 1995, Stiglitz 2003). A combination of technological change and a move in the political centre to the Right have led to increased economic and cultural globalisation, with, for example, the privatisation of utilities and services, reductions in the power of unions, increased freedoms regarding movement of capital, and a shift in attitudes regarding taxation (Hutton 1996). The recession of the early 1980s in the UK began a ‘hollowing out’ of mid-range jobs in the UK (Fothergill et al 1985), with many geographical areas dependent on manufacturing (including mining and shipbuilding), as well as areas dependent on such industries and the wages of those workers (for example, a large number of coastal holiday towns), propelled into a process of decline from which they have never re-emerged. This process gathered pace in the following decades to produce what researchers now describe as the ‘hourglass economy’, where employment is becoming polarised with increasing numbers of ‘high quality’ and ‘low quality’ jobs, but a continuing decline in routine occupations in the middle of the labour market (Sissons 2011). As Goos and Manning describe, the UK economy is becoming divided into ‘lovely’ and ‘lousy’ jobs (Goos and Manning 2007, Goos et al 2010).

Technology has displaced routine occupations such as secretarial and administrative work previously done by women, for example, and machine-based factory work, assembly line operations and so on for men (Sissons 2011). There has been some growth in knowledge-intensive work, which implies a greater need for those who understand how technology works, and in particular how to use and control it. One particular consequence has been the growth in the need for managers of seemingly a range of varieties. There are also a large number of low-skilled jobs whose existence (and even growth) may not be attributable to technology, but neither can technology readily replace them – occupations such as those in social care, construction and personal services have grown significantly in the last 30 years (Sissons 2011).

The work of Goos and Manning casts further doubt on the idea that technology will produce more knowledge-intensive jobs and lead to the withering away of low-skilled work. Goos and Manning (2007) argue that technology may be driving an increase in demand for (p.111) those in more high-skilled jobs, but it is the increased supply of those willing to do low-skilled jobs who have been displaced from routine occupations that is creating the conditions for the ‘hourglass economy’.

Work by Michaels et al (2014), looking at the role of technology in increasing polarisation in 11 different countries, supports this view, but only to an extent. They argue that it accounts for only up to a quarter of the growth in demand for more skilled workers, and so there must be other factors at play here. Their research suggests, for example, that the growth in some areas of low-skilled work (such as personal services) is a function of increases in demand from high earners for such services as well as increased female participation in the labour market.

The impact on wages of this occupational polarisation appears more ambiguous. Technology is mediated to a greater extent in its impact on wages by factors such as unionisation, differences between and within occupations, gender and also the availability of workers able to do what employers want (Holmes 2010). Hence, research by Holmes and Mayhew (2012) shows that the relative wages of those in these routine occupations may have held up, even though their numbers have decreased. There is also evidence to suggest that the wages of the new low-skilled workers have also held up, relative to more middle earners, despite the supply of these workers increasing (Holmes 2010), which may be due to the impact of the minimum wage legislation in the UK. This legislation has only been partially effective, though. Since the late 2000s, the stagnation in real wages in the UK economy has had a particularly serious impact on those at the low end of the wages ladder. The Resolution Foundation, a charity dedicated to looking at low pay, argues that ‘since 2009 in particular, the number of workers earnings less than a living wage – the amount considered adequate to achieve a minimum standard of living – has rocketed, from 3.4 million to 4.8 million in April 2012’ (Whittaker and Hurrell 2013: 4).

The evidence also shows that for a significant proportion of such workers, there is no route out of these occupations, which has significant implications for social mobility. Sissons’ analysis of earnings mobility in the 2000s, using the British Household Panel Survey, showed that 60% of those in the bottom 10% of the earning distribution in 2001 had remained within the bottom three deciles by 2007. The Resolution Foundation describes the increase in those in low-pay situations in the UK as a ‘structural one’ (Whittaker and Hurrell 2013: 22), stemming from an increasing demand for low-paid service work resulting from changes in technology and patterns of consumption. Those receiving less than two-thirds of the average median living wage per hour, 47% (p.112) are aged 31-60. Low-paid work is not confined to the young and old, however; it is prevalent across the lifecourse, as Figure 6.1 shows.

The shape of the labour market: hourglass, diamond or molecule?

Figure 6.1: Distribution of low pay across age groups, 1975–2011

Source: Whittaker and Hurrell (2013)

Sizeable numbers of low-qualified workers are stuck in occupations that are less likely to be skilled. The disconnect in the 21st-century labour market between low-skilled and mid-range work makes the possibilities of upward mobility far lResearch by Nunn et al (2007) and Holmes (2010) reach similar conclusions regarding the solidifying of divisions between the low- and high-skilled. There may be a higher chance of workers when they enter low-skilled work remaining there throughout their careers, as the ‘small rungs’ in the career ladder that used to be available to them are no longer there. There may also be a continued process of what Sissons describes as ‘bumping down’, where routine workers, as their occupational areas shrink, come to swell the ranks of the non-routine, low-skilled workers in the labour market ‘tier’ below.

The shape and size of the hourglass(es)

While the aggregate analysis seems to show an increasing polarisation between high-skilled/high paid work and low-skilled/low paid work, there are significant variations in the nature, extent and shape of this divide by factors such as gender, geography and occupation.

Sissons illustrates how, while polarisation is affecting both men and women, it is doing so in very different ways, mainly because of the very strong gendering of the labour market that still exists. The fastest-growing (p.113) jobs over 2001 to 2007 for men and women were ones that reflected this gendering exceptionally strongly. For men, there was a growth in over 120,000 in employment in construction, and for women, a growth in over 180,000 in childcare and related services. For all the rhetoric of a ‘new’ knowledge-based economy, the evidence appears to show that the old economy, structured around low-skilled, gender-differentiated work, is alive and well.

Differences within the labour market by geography are equally, if not more, striking. In 2012 Green undertook research to explore the utilisation of skills in three different parts of England and Wales as part of a four-country study into labour market structures at the local level. Green argued that:

It was clear that there is strong variation between local labour markets in terms of the relationship between skills supply and demand. While some sub-regions exhibit characteristics of ‘high skills equilibrium’, with a strong supply of skills being matched by a strong demand for skills, others experience either an imbalance in skills supply and demand (leading to ‘skills shortages’ or ‘skills surplus’) or a ‘low skills equilibrium’ where low skills in the workforce are matched by low demand for skills amongst employers.

(Green 2012: 10)

There is significant research in the UK showing the growing differences between regional economies and labour markets (Gardiner et al 2012). Neither is it clear the type of local labour market that would favour economic mobility for the low-skilled are the ones we would expect. It may be that those labour markets that retain a presence from the type of mid-range occupations threatened by technology may benefit this group more than the more ‘successful’ economies, as the former can provide conventional ladders out of low-skilled work. Conversely, the labour market of the UK’s most knowledge-rich regional economy, London, may provide less opportunity. It is the most extreme form of hourglass – the gaps between low and high skill are more extreme, and the career ladder’ is steeper and more difficult to ascend (Kaplanis 2007, Deloitte 2013).

As Green acknowledges, though, even the framework she outlines is a simplistic one. The boundaries of what a ‘local labour market’ constitutes are contested and contingent. Locality in labour market terms means different things to different people, often because of their differing existing skill levels and the occupational field they are in. (p.114) Most UK-born workers in the UK are still relatively geographically immobile (it is quite a different story for migrant labour; Green et al 2013). However, this differs as one moves up the skill ladder (Holmes 2010). For more highly skilled workers, their labour market is national or international. Moreover, it then differs again by occupational field. As evidence from both the UK and Europe shows for those in certain high-demand/high-skilled occupations, they operate in a global ‘local labour market’ (Cedefop 2011).

Nor is the picture the same for workers of different generations. Holmes and Tholen (2013) examined the labour market experiences of those who entered the labour market in the mid-1970s and the late 1980s respectively. They found that, for the younger cohort, after the recession of the late 2000s it was harder to move up through the occupational scale. Something was happening that was changing the nature of progression: ‘The analysis from this paper suggests that in recent years, increased room at the top has not increased upward mobility for those already in work, implying that a growing number of these jobs are predominantly recruiting new labour market entrants rather than offering opportunities’ (Holmes and Tholen 2013: 38). However, this pattern was not the same across occupations, meaning that depending on the field one worked in, the prospects looked better or worse.

While it is an appealing image in terms of its ability to make a complex thing like the labour market look simple in reality, the labour market is not really one hourglass. If anything, it is thousands of smaller hourglasses, where the relative size of the bottom, middle and top differs. And in some occupational or geographical areas, it may not resemble an hourglass at all. Occupational mobility is a complex phenomenon, experienced and lived through gender, ethnicity, region and occupation. In this context, it becomes clear that the study of it must reflect these nuances and how they interact at the individual level. This complexity also suggests that success and progress may mean very different things across individuals, groups and, in particular, localities. The fixation, among many politicians, with long-range upward occupational mobility means that the majority of occupational social mobility is being lost from political view, precisely because of these nuances. This is a contradictory outcome. Politicians understand acutely the importance of locality in shaping people’s lives; their careers depend on this understanding. Yet where social mobility is concerned, they fail to make this link, preferring to fall back to the familiar celebration of long-range mobility where they find it.

(p.115) The need for low-skilled work

While the shape of the hourglass may differ by region, occupation and so on, what appears certain is that there will not be a disappearance of what is presently defined as lower-skilled/lower-paid work in the foreseeable future. And this reality poses major questions for politicians. It is unrealistic to suggest that all workers will be able to benefit from any set of social mobility policies whose main aim is to facilitate upwards movement from low-skilled to higher-skilled work. Neither is it just an issue of the pressure of numbers implying that people have to remain in these jobs as there is no room for them in the other parts of the hourglass. While the desirability of upward economic mobility is taken as given at present, there will be people who do not want to move out of their social and economic locations.

As Friedman (2013) has argued, many people do not necessarily wish to move jobs; rather, they want their existing ones to improve. However, this does not mean they do not want to see progress in their working lives in the context of these jobs. Increasing pay is one obvious, essential route, to such progress of course. There are increasing numbers in poorly paid work in the UK, and a rising amount of in-work poverty. This was picked up on by the Labour Party in the early 2010s, coining the phrase ‘pre-distribution’ to describe the need to get employers to pay better wages to the low-skilled (Labour Party 2013).

As crucial as increasing wages is, though, it is not enough. US economist Richard Florida argued in 2010 that low-skilled work should be ‘upgraded’. Florida looks back at the routine ‘blue-collar’ jobs, which have been lost from the middle of the occupational range. He points to how many of these jobs are now viewed with great affection for their security and status, even though they were once seen as ‘bad’ jobs:

… more than 60 million Americans toil in low-wage, low-skill service jobs in everything from food prep and retail sales to personal care. We can transform them into good, family sustaining jobs, the same way we made manufacturing jobs good jobs decades ago, by creatifying them – tapping the knowledge and creativity of workers as a source of productivity, which in turn will generate higher wages.

(quoted in Business Insider 2012: 2)

This process of ‘creatifying’ involves tapping into the skills and knowledge that each individual has, and creating structures inside (p.116) and outside the workplace that enable them to be maximised. While upgrading work might be only one way of describing it, the idea that something more could and should be done to enable those in low-paid positions to experience upward social mobility is important. Improving earnings is crucial – economic progression is at the centre of any understanding of mobility, including the one being developed in this book. Improving earnings also increases status and contributes to the perception of the job among those who do not do it (although, as will be argued later, this is problematic). But the quality and nature of the job is also vital. Increasing wages for low-skilled occupations on its own does not constitute a way of ensuring that those occupying these positions are socially mobile. The nature of the work undertaken must also be examined. The strengths of the blue-collar jobs described by Florida were founded on more than income. There were relatively high levels of job satisfaction, often rooted in the solidarity of the working experience. And recreating these conditions in the 21st century will be very difficult. As the research on job polarisation shows, the large workplaces built on routine labour are less common. Where they do exist, the fragmentation of social groupings among the workforce by gender and ethnicity makes the kind of connection between work and community on which these jobs were based far less relevant. There is also a high degree of mythologising here. The blue-collar idyll was a predominantly male, white one, and the work still routine and stultifying (Goldthorpe et al 1968, Ramdin 2007).

Moreover, it was the lack of progression afforded by this work that made so many skilled working-class families keen to ensure that their children had something better in the last quarter of the 20th century in the UK, driving up HE participation rates and helping to embed aspiration more firmly in the national value system (Goldthorpe et al 1968). If it was a lack of social mobility that was the problem for this kind of work, it may seem perverse to see it as part of the solution to social mobility problems today. Looking back, then, is only a partial remedy for the problem of social mobility and low-skilled work. Thinking about why blue-collar work was seen as good, though, highlights the importance of non-pay factors. Is it possible to experience progress in a job if you remain in the same job and your pay remains relatively unchanged (and low)?

For this to happen, both the perception and experience of lower-skilled work would have to change. There are examples of how this could be done. In 2009 nef undertook research into the social value of six different occupations. The additional social value of those working in childcare was up to £9.50 for every £1 spent, in contrast (p.117) to bankers, who may collect salaries in excess of £500,000, but destroy £7 of social value for every £1 they generate. Measuring social value as opposed to pay is a different way of also capturing success where employment is concerned. If social value could be articulated as a metric of achievement, then it would greatly support attempts to broaden what social mobility means (nef 2009).

On a more practical level, Anderton and Bevan’s research for the Work Foundation in 2014 examined how employers could foster greater employee engagement and involvement to enrich what they described as ‘constrained jobs’, where worker autonomy is limited and the kind of ‘digital Taylorism’ described in Chapter Two has taken hold. They list four different ways by which employers could do this:

  • New performance management: shifting from traditional object and quantitative to qualitative performance measures.

  • An emphasis on training, development, problem solving and career progression.

  • Team work and participation in business development.

  • Flexible working and supporting work–home balance (Anderton and Bevan 2014:7)

Upgrading lower-skilled work is not going to be easy. The UK has a relatively high number of jobs that are in ‘low-skilled categories’ (Holmes 2010).

Over 20% of workers are in jobs that require only a primary level of education. This compares to less than 10% in Sweden and Germany (Osborne 2014). And low-skilled work is overrepresented in the jobs that are rated by employees as the least satisfying (Addley 2014).

Finally, attempts that have been undertaken to upgrade occupational areas have also been controversial. The recent change in nursing in England to become a graduate occupation is an instructive example. In 2009, measures were put in place to make nursing a graduate profession by 2013 (BBC News 2009). All new nurses now need a graduate-level qualification. This was opposed by some, who felt that the academic knowledge associated with graduate study was not required for good nursing, and that it was placing new barriers in front of those who wanted to become nurses (Ford 2009). The extent to which nursing practice has been improved by this change is no doubt contested. It shows, however, that it is possible for policy makers to engineer upward economic social mobility by shifting the required skill levels of a whole group of workers.

(p.118) The nature of work’

Exploring the social mobility challenges facing low-skilled workers is also important because of the wider set of questions regarding the relationship between social mobility and the nature of work it opens up. It reaffirms, for example, the importance of intra- as well as intergenerational mobility. Considering what happens when there is likely to be occupational immobility (in terms of socioeconomic classification) brings into greater focus work over the lifecourse. The concept of a career trajectory is lost when the attention remains predominantly on intergenerational mobility. The ideas of permanent income and occupational maturity described in Chapter Two may be good and accurate ways of summarising earnings or occupation over the lifecourse, but they tell us nothing about the lifecourse itself. This suggests the need for different ways of studying social mobility that may draw on qualitative as well as quantitative forms of analysis. Life history research, for example, can tell us as much about social mobility as any of the quantitative studies that have come to define the nature of study in the field (Stuart 2012).

The limitations in using the movement between jobs as a measure of social mobility are not confined to the case of the low-skilled. They are not the only group who may remain in their jobs for a considerable period of time, or within the same occupational group. While the level of job change is higher in the 21st century (Cabinet Office 2009), it still appears to be the case that the majority of workers remain in fewer than 10 jobs over their career, and job change is skewed towards the early years, when career paths are being established (Bialik 2010, National Careers Council 2013).

What happens within jobs matters as much for the high- and mid-skilled workers as for the low-skilled. Restricting a measure of mobility over the lifecourse to changes in occupation doesn’t capture what happens within the experience of a particular job. This can include significant increases in pay and responsibility, even though the job title doesn’t change (or decreases in pay and responsibility or demotion as well). Deciding what metrics could be used to capture progression within jobs or occupations is very difficult, though. There is significant evidence to support the view that ‘job satisfaction’ matters, and not just to what happens in work time but also to an individual’s overall view on their well-being or progress in life (Clark 1998, 2005; Diaz-Serrano and Cabral Vieira 2005). But there is far less consensus on what ‘job satisfaction’ means, and whether it is so amenable to fluctuation, dependent on when it is measured (today I may feel satisfied with my (p.119) job as I had a good day, but tomorrow I may lose a deal and feel unhappy that you might not be able to measure it at all (van Saane et al 2003).

More sophisticated attempts to capture job satisfaction exist, which concentrate on the more ongoing features of the working experience that may underpin people’s long-term views on their work (Rose 2003). Improving these more permanent features of the working experience provides a potentially more robust approach to measuring the occupational part of social mobility than merely asking workers whether they are happy or not. However, the degree to which there are practices firmly in place that can deliver this kind of progression in the UK has to be questioned. A major part of the ‘education sceptics’’ argument described in the last chapter was that too many workers in the UK are underemployed. Keep and James (2010) argue that the structures are not in place to utilise and develop the skills of the workforce in the UK.

This view is supported by the UKCES (2011) and its work on high performance working (HPW). HPW touches on some of the themes described by Anderson and Bevan (2014) in their research on low-skilled occupations. It is a broad concept, summarised by Belt and Giles as:

… a general approach to managing organisations that aims to stimulate more effective employee involvement and commitment in order to achieve high levels of performance. The precise form HPW takes within an organisation will vary depending on context, but will include activities in the areas of: human resource management (eg pay and incentives, appraisal, workforce development), work organisation (eg team working and job design), employment relations, management and leadership (including strategic management and business development as well as line management), and organisational development. Importantly, the HPW approach is specifically designed to enhance the discretionary effort employees put into their work, and to fully utilise the skills that they possess. It needs to be underpinned by a philosophy of people management that emphasises autonomy, participation and learning.

(Belt and Giles 2009: 10)

UKCES argues that there is a significant ‘policy to implementation’ gap where HPW is concerned. It is not well understood by employers or, for that matter, by policy makers – the incentives to drive its (p.120) development are too voluntary. As Belt and Giles 2009:10) state it is a ‘policy without a home’ in the UK.

‘The meaning of work

Even if the development of HPW could be incentivised and more clearly defined, there are still some wider questions about work that need to be addressed.

A holistic approach is required here, as examining/promoting social mobility should involve not only improving the experience of work, but also asking what it is for and what it does – or should – mean. This requires simultaneously paying more and less attention to work. Keynes predicted that as societies became richer, the time spent in work would decline, and we would move to a more leisure-based society (Keynes 1931; see also Veal 2009). Yet this has not been the case. In the UK, in particular, there has been a move in the other direction, with longer working hours, and an increasing blurring for those in higher-skilled occupations between work and non-work life. This has been described as the ‘work–life’ merge, where technology enables people to be constantly available for work to answer emails and so on (Hinsliff 2013). In France, however, they have recognised the impact of this work–life merge on employees and introduced legislation to curtail email communication out of work hours (de Castella 2014).

The work–life merge phenomenon is not entirely employer-driven. Work by Gershuny and Fisher in 2014 showed that (in contrast to the mid to late 20th century) in the early 21st century, the highest skilled and highest paid workers want to work more, as they find this both rewarding and where they get meaning in life. There has been the growth of a ‘super-ordinate working class’, which cuts across both genders, who are reversing historical trends by working more than those in lower-skilled/lower-earning groups (Gershuny and Fisher 2014).

The rise of the work–life merge and the super-ordinate working class has an importance that spreads wider than just this super-ordinate elite to affect how the rest of society work and see work. It is shaping how success is understood, and can filter down to influence attitudes and practices among those not paid as handsomely for answering their emails in the middle of the night. Across the workforce this kind of commitment can easily become an expectation. What some people are doing because they choose to, others then do because they have to. (This is indeed if the super-ordinate class are really engaged in the work–life merge because they want to, or because they are caught in a form of executive ‘prisoner’s dilemma’, where everyone else is doing (p.121) it, so they feel they have to too, in order to retain their position.) The rise of the super-ordinate working class may be the latest facet of what Sennett described in his 1999 book, The Corrosion of Character as the ‘new economy’. Sennett argued that new opportunities for self-fulfilment in work, represented by technology and new ways of working, are actually a chimera; in reality, greater flexibility in the labour market and workplace only results in new forms of oppression.


This chapter set out to take a more detailed look at how contemporary changes in the labour market that have gathered pace since the 1970s have shaped, and been shaped by, how social mobility is understood. While the hourglass is a good visual metaphor for the labour market in an economy like the UK’s in the early 21st century, no two hourglasses ever look the same, and neither do any two parts of the labour market. The drawback with any metaphor designed to make complex issues easier to understand is that it may mask the details that make up the real picture. In reality, the UK economy is more complex. The hourglass is a fairly one-dimensional model, where there is a clearly defined top, middle and bottom. This chapter has tried to show that the picture may be more nuanced than this, and the reality may be more three-dimensional than one-dimensional. The relationships between ‘top’, ‘middle’ and ‘bottom’ might actually start to be less clear-cut where work is concerned.

There are alternative metaphors for how the labour market currently looks. First, it may look more like a pyramid shape where there are those at the top whose wealth and power allow them to insulate themselves from any of the negative side effects of high-paid work. At the bottom there could be a growing group whose work is insecure, low paid and lacking satisfaction. The differentiator of the degree, for example, in deciding who is in the top half of the hourglass or the bottom, may (if the evidence in Chapter One is accepted) mean less. The title of the job, which is the basis by which researchers place people in the hourglass model, may also mean less than we think, and it might require more detailed qualitative work to really decide who is – or who is not – going up or down the occupational scale. As Lloyd and Payne argue, it is too easily assumed that we know what skill means. At present there is an ‘often un-stated, question: what is “skill” and what do we mean by a “skilled job?”’ (Lloyd and Payne 2008: 1).

An even better metaphor may be many molecules. Each has its own individual structure, and the atoms within are connected not by (p.122) a straight line from top to bottom, but by a number of different lines: vertical, horizontal and sloping, connecting separate, distinct parts of the whole. In this more three-dimensional conception of the labour market, it is more difficult to say what sort of mobility (downward, upward or sideways) that movement between jobs represents, and how different jobs relate to each other. The appeal of this metaphor is that it at least attempts to reflect the complexities described, the more complicated nature of work as provider of income, status, meaning and social mobility, and it fits the idea of social mobility as a more complex phenomenon. It attempts to reflect the need to explore not just progression within or between jobs in terms of income or status, but also in the nature of what the work itself means to the worker. Increases in income and responsibility may be all well and good, but beyond a certain level, do they really constitute progress or success?