Abstract
ABSTRACT
There is currently widespread concern that Britain’s cultural and creative industries (CCIs) are increasingly dominated by the privileged. This stands in stark contrast to dominant policy narratives of the CCIs as meritocratic. Until now this debate has been clouded by a relative paucity of data on class origins. This paper draws on new social origin data from the 2014 Labour Force Survey to provide the first large-scale, representative study of the class composition of Britain’s creative workforce. The analysis demonstrates that CCIs show significant variation in their individual “openness”, although there is a general under-representation of those from working-class origins across the sector. This under-representation is especially pronounced in publishing and music, in contrast to, for example, craft. Moreover, even when those from working-class backgrounds enter certain CCIs, they face a “class origin pay gap” compared to those from privileged backgrounds. The paper discusses how class inequalities, as well as those related to gender and ethnicity, between individual CCIs point to occupational subcultures that resist aggregation into the Department for Culture, Media and Sport’s broader category of CCIs. The paper concludes by suggesting the importance of disaggregating CCIs and rethinking the definition and boundaries of CCIs as a meaningful category.
Introduction
There is currently widespread concern that Britain’s cultural and creative industries (CCIs), contrary to their image as emblems of a new, fluid and dynamic, “knowledge economy”, are increasingly dominated by those from privileged class origins (Hough, 2012; Plunkett, 2014). Public policy is beginning to take notice of these issues, with the UK’s Arts Minister Ed Vaizey (2016) pledging to support a range of initiatives aimed at addressing issues of diversity in cultural production. Moreover, this concern is set against the backdrop of transatlantic concerns over Black, Asian, and Minority Ethnic (BAME) representation in cultural production (Revior, 2016).
At present, however, the debates have generated more heat than light, with concerns over diversity often expressed (or simply dismissed) by way of individual anecdotes about the career successes and struggles of particular high-profile cultural producers (Sherwin, 2016). Attempts to ground these debates in more empirical work have been hampered by a lack of appropriate data. For those interested in the representation of different class groups within the CCI workforce this can be a particular problem. The most detailed data on the composition of the creative workforce, for example provided by Creative Skillset (2012) or, most recently by the UK’s Department for Culture, Media and Sport (DCMS, 2015b), lacks any information about social origin. This paper therefore capitalises on data from the 2014 Labour Force Survey (LFS) to provide the first large-scale, representative study of the class composition of Britain’s cultural workforce. In doing so, it returns to discussions of the meaning, definition, and coherence of CCIs as a sector.
In this study, we make three interventions. First, we reject policy narratives portraying the CCIs as open and meritocratic, demonstrating that those from working-class backgrounds are significantly under-represented within the CCIs. Investigating this further, our analysis shows significant variation in the “openness” of individual CCI sectors, and here we can note in particular the social exclusivity of publishing and music and the relative openness of crafts. Second, it shows that even when those from working-class backgrounds enter certain CCIs, such as museums, galleries, libraries, and IT, they face a “class origin pay gap” compared to those from higher professional and managerial backgrounds. Third, reflecting on these distinct class inequalities in access and pay between individual CCIs, we underline the existence of particular occupational subcultures that resist aggregation into DCMS’ broader category of CCIs. This latter point raises questions as to the very coherence of CCIs and therefore problematises future uses of this aggregate category.
The paper proceeds as follows. It begins with a short summary of the literature surrounding the nature of work within CCIs. Here, we consider two contrasting, if not contradictory, themes at the heart of the CCI narrative: first, that this occupational sector represents an open, meritocratic blueprint for a new form of post-industrial economy; and second that working conditions within CCIs are poor, with uncertain career prospects and low- or even no-pay. The former view is most closely associated with think tank and policy discourses, while the latter has led academics interested in the structural conditions of CCI labour to question the meritocracy narrative altogether.
The analysis that follows this discussion directly addresses these two positions in turn, refuting the first and offering a new empirical basis from which to assess the claims associated with the second. Here we begin by exploring the employment profiles of the CCIs in the LFS data using descriptive statistics before turning to regression analysis in order to highlight the inequalities within the UK’s CCI sector.
Taken together, these approaches indicate clear differences between the occupations that are currently aggregated together by policy-makers to represent the economic performance of CCIs. The article concludes by linking the differences in occupational subcultures to the long running debate over how to define CCIs. These differences suggest that, by continuing to aggregate very different occupations, CCI policy will inevitably be inappropriate for all of the sectors included within DCMS’ current categorisation. The analysis suggests, therefore, that the question of what, if anything, binds these sectors together should be reopened, given the fact that the aggregate category of CCIs cannot be based on similarities of occupational structure and culture.
Creativity, meritocracy, and the CCIs
The CCIs, in the UK and elsewhere, have long been subject to claims about their economic potential. These claims have been rooted in how CCIs have been defined as a sector of the economy that can outperform other types of occupations, and – our substantive focus here – their distinctively open and meritocratic nature. Florida’s 2002 work – which we take up further below – is the canonical, if much critiqued, text on this point.
The current definition of CCIs uses the idea of creativity as the basis for claims about CCIs’ economic potential. Creativity has a dual role within the definition, linking seemingly diverse occupations, including the arts, software professionals and media work, as well as providing the basis for analysis of economic survey data. The most recent sets of these figures (DCMS, 2014, 2015a, 2015b) are based on Creative Skillset’s idea of “creative intensity”, whereby:
in essence a creative industry is defined as being one which employs a significant proportion of creative people, as identified by those being employed in a creative occupation. (Bakshi, Freeman, & Higgs, 2013, p. 11)
And creativity is:
A role within the creative process that brings cognitive skills to bear to bring about a differentiation to yield either novel, or significantly enhanced products whose final form is not fully specified in advance. (Bakshi, Freeman, & Higgs, 2013, p. 24)
The idea of creativity as the basis for occupational aggregation yields nine occupational clusters: Advertising and marketing; Architecture; Crafts; Design; Film, TV, video, radio, and photography; IT, software, and computer services; Publishing; Museums, galleries and libraries; Music, performing and visual arts. These occupations are closely related to a specific vision of work in a post-industrial economy, whereby work is centred on making a living based on citizens’ capacity as creative individuals. The citizens’ role is not to produce material goods, such as cars, washing machines, or foodstuffs. Rather the immaterial products associated with services and intellectual property are at the heart of work in this version of the modern British economy.
As McRobbie (2015) has suggested, this is simultaneously a narrative that promotes an idea of the culture of work in the CCIs as one in which creativity is located as a central mechanism of la carrière ouverte aux talents. In other words, the “creative” job is supposedly open to everyone. In British policy and practice discourses, where citizens are rarely described as anything other than innately creative (O’Brien, 2014), it is then a short step, within policy and practices discourses, to suggest that those who are able to make a living by capitalising on their creativity are simply reaping the just desserts of talent and skill. Creative work can, therefore, be read as intertwined with ideas of meritocracy prevailing broadly across modern economic and social organisation (Littler, 2013; O’Brien, 2014).
The most powerful account of the meritocratic character of creative work was famously provided by Richard Florida, in his account of the rise of the “creative class” (2002). Florida argued that the economic advantage of cities in the “new age of creativity” is increasingly bound up with their ability to attract a cohort of young, highly educated workers with specialist forms of creative, technical knowledge. Crucially, this involves facilitating an environment that aligns with what Florida identifies as this group’s core characteristics; namely, its diverse origins, social connectivity, cultural eclecticism, and meritocratic ethos (see O’Brien, 2014; Miles, 2016 for critiques of these characteristics). Following in the wake of Florida’s analysis, various think tanks and policy documents have narrated the CCIs as a dynamic, highly skill-based, sector of the economy, especially symbolic of meritocratic recruitment and working practices, which are in turn considered to be crucial to the sector’s success.
Sociological and cultural studies accounts of the creative industries have subjected these claims to detailed criticism. Most notably, scholars have questioned both the working conditions found in CCIs (Hesmondhalgh & Baker, 2010; McRobbie, 2002), and also the narratives of meritocracy attached to those who work in these sectors of the economy. In the United States, Koppman’s (2015) work has shown how shared cultural tastes correlated with middle class backgrounds are highly influential in hiring practices within CCIs, concurring with Rivera (2015) that hiring is, in effect, a form of cultural matching rather than a meritocratic exercise. Moreover this meritocratic narrative serves to obscure structural inequalities associated with gender (Gill, 2002), class (Friedman, O’Brien, & Laurison, 2016) and other forms of discrimination (Littler, 2013).
The notion that the UK creative occupational field is socially open or meritocratic is also challenged by more contemporary accounts of its diversity and accessibility from within cultural policy studies (e.g. Allen, Quinn, Hollingworth, & Rose, 2010). This research highlights inequalities associated with gender and ethnicity, with a recent review of the literature by O’Brien and Oakley (2016) demonstrating structural inequalities resulting from organisational issues, work patterns, hiring practices, and – a central focus in this paper – discriminatory pay gaps (Creative Skillset, 2012).
Work examining the role of class inequality within the CCIs is markedly less developed. As highlighted recently by O’Brien and Oakley (2016), this is in large part due to the lack of large-scale representative data documenting the class origins of those working in the CCIs. There is, however, an important body of work that probes the way class connects to occupational access in specific CCIs. This has focused on the classed nature of particular educational pathways (e.g. Allen, 2014; Banks & Oakley, 2015; Bull, 2014; Scharff, 2015) the way the privileged often draw upon powerful social networks in forging cultural careers (Grugulis & Stoyanova, 2012; Nelligan, 2015), or the significant barriers to entry faced by those from working-class backgrounds attempting to move into the CCIs (Eikhof & Warhurst, 2012; Friedman et al., 2016; Randle, Forson, & Calveley, 2015).
Even more recently, Miles (2016) has drawn on the Great British Class Survey (GBCS) to provide arguably the most detailed understanding of the social composition of Britain’s cultural occupations. Drawing on the work of Grusky and his various collaborators (e.g. Grusky & Sorensen, 1998; Grusky & Weeden, 2008), Miles explores the “micro class” dimensions of occupations in the creative sector, using GBCS data to compare their recruitment profiles, assets (in terms of economic, social, and cultural capital) and values. His findings suggest that there are considerable variations in experience, resource, and outlook, even between the occupational groups that Florida includes in his “super-creative core”. This point is in keeping with longstanding debates over the definition and demarcation of CCIs (Campbell, 2014), a matter with which this paper engages in its penultimate section.
This paper seeks to extend Miles’ work in two key empirical directions. First, by drawing on the LFS, we provide the first nationally representative picture of the class composition of Britain’s CCIs – as well as how this relates to inequalities of gender, ethnicity, and education. Second, drawing on the feminist concept of the “glass ceiling”, we look at earnings of employees within the CCIs and how these may be affected by class origin. For example, recent work (Laurison & Friedman, 2016) has identified that even when those from working-class backgrounds are upwardly mobile into Britain’s high-status occupations they face a “class origin pay gap” that prevents them from enjoying equivalent earnings to those from intergenerationally stable backgrounds. More specifically, they find that those whose parents were employed in semi-routine and routine (NS-SEC 6–8) occupations earn on average £7000 less than colleagues from higher professional and managerial backgrounds – even after controlling for a host of factors known to affect earnings. Here we therefore explore whether this finding of a “class ceiling” also obtains in cultural work.
Data and methods
As noted, we draw here on data from the Office of National Statistics’ quarterly LFS, specifically data pooled from four quarterly surveys from October 2013 to September 2014. We first used the DCMS Creative Industries Estimates (DCMS, 2015a) to assign occupations (based on 4-digit SOC2010 codes) to nine sectors of the CCIs. The 30 individual occupations in each of these sectors are listed in Table 3, with a total of 2201 respondents working in these occupations when they responded to the survey. We then identified the respondents employed in these occupations who also responded to the social origin question in the July–September 2014 survey (1769 respondents). This question asks respondents the occupation of the main earner parent when they were 14. We then group respondents’ social origin into four groups based on the National Statistics Socio-economic Classification (NS-SEC) classes; those with parents in NS-SEC 1 (higher professional and managerial occupations), in NS-SEC 2 (lower professional and managerial positions), NS-SEC 3, 4, or 5 (intermediate occupations or self-employed), or NS-SEC 6–8 (semi-routine, routine occupations, or unemployed). We also removed all those under 23,1 in full-time education, or over 69, as the LFS collects data on those over 69 differently, since most people in this age group have moved into retirement. This leaves 1637 respondents in CCI occupations, and 918 who also have earnings information (862 with data on all covariates used in regression models).
It is important to note that the LFS does not collect earnings information for respondents who are self-employed; thus all reports of earnings below are only for those who are employees. The self-employed are included in our descriptive statistics below, but we are unable to say anything here about the situation for self-employed workers in the CCIs. Appendix table A1 shows the proportion of workers in each sector in each NS-SEC category. Finally the analyses use the recommended survey weighting from the LFS in all analyses, but were replicated with no weighting and the results were found to be consistent; full descriptions of variables and other methodological notes are in the appendix.
Understanding the creative workforce: evidence from 2014 LFS
Who are the creative workers?
We begin our analysis with a descriptive portrait of the demographic composition of the different sectors that make up the CCIs in the UK (according to DCMS’s definition). Table 1 thus reports the relative size of each sector among LFS respondents, their gender and ethnic makeup, and the percentage of workers with degrees. Figure 1 also reports average weekly earnings of employees in each sector.
Are the creative industries meritocratic? An analysis of the 2014 British Labour Force Survey
Published online:
12 April 2016Figure 1. Average weekly earnings in the CCIs, by sector. Note: Average weekly earnings for all non-self-employed respondents reporting an occupation assigned to a CCI sector, aged 23–69 and not in full-time education.
Figure 1. Average weekly earnings in the CCIs, by sector. Note: Average weekly earnings for all non-self-employed respondents reporting an occupation assigned to a CCI sector, aged 23–69 and not in full-time education.
Are the creative industries meritocratic? An analysis of the 2014 British Labour Force Survey
Published online:
12 April 2016Table 1. Composition of the culture and creative industries.
Table 1 and Figure 1 point to three significant findings. First, in terms of economic contribution, Table 1 demonstrates that IT, Software, and Computer Services is by far the biggest employer, followed by advertising. Moreover, Figure 1 illustrates that these two sectors also have the highest average earnings within the sector, nearly £100/week more than the average for the CCIs as a whole.
Second, Table 1 shows familiar demographic skews within the CCIs. In line with the recent DCMS (2015b) report, our analysis demonstrates that women are significantly underrepresented in the CCIs, BAME groups are marginally underrepresented, and all workers are significantly better-educated than the population as a whole.
However, these aggregate figures hide significant variations by individual sector. For example, only the IT sector has a higher percentage of BAME employees than the general population, while every other creative occupation is more white than the UK as a whole. There is also particularly acute under-representation of women in architecture, craft, film and TV, and IT.
Third, Figure 1 demonstrates that earnings for employees (as distinct from all workers) within the CCIs are much higher than the population as a whole. While much recent debate has focused on the precariousness and low pay of cultural labour (Hesmondlagh & Baker, 2010; McRobbie, 2002), Figure 1 suggests that the CCIs actually offer rates of pay close to that of higher managers and professionals (NS-SEC 1). Thus, the weekly average earnings for those in the CCIs is £801, compared to £896 for those in NS-SEC 1, £582 for those in NS-SEC 2, and £522 in the workforce as a whole. This is partially explained by the fact that 33% of people in CCI occupations are themselves classed as higher managerial and professional, but even when these are excluded employees in the CCI sector are still comparatively better paid than those in the labour force as a whole (the average pay of people in CCIs outside NS-SEC 1 is £613/week).
Again, though, there is the need to re-state that these data only cover employees, suggesting the need for the continued narrative of insecurity and low pay for those who do not manage to secure stable employment in the sector. In addition, the headline figures mask important inter-sector differences. Occupations associated with Music, Museums, and the performing arts – where indeed a lot of the qualitative research on precarious labour has emerged (Banks, Gill, & Taylor, 2014) – have markedly low average rates of pay closer to that of intermediate and routine occupations.
The class origins of the cultural workforce
While results so far echo relatively well-documented demographic and earnings inequalities within the CCIs, very little is known about how these map onto the class origins of those employed in the cultural sector. Table 2 therefore examines the social origins of those employed in the CCIs as a whole, and then shows how these compare to the origins of those in higher professional and managerial occupations (NS-SEC 1) and lower professional and managerial occupations (NS-SEC 2), in the population as a whole. Despite the dominant policy narratives of openness and meritocracy, Table 2 shows clearly that there is a significant under-representation of people from working-class origins in creative occupations. While 34.7% of the UK population aged 23–69 had a parent employed in a routine or semi-routine working-class occupation, the figure among those working in the CCIs is only 18%. This under-representation is mirrored by the comparative over-representation of those from professional and managerial backgrounds (that is, NS-SEC 1 and 2 combined: 50% in the CCIs vs. 29.1% in the population). It is also telling that the CCI skew towards those from privileged backgrounds closely mirrors that of Britain’s highest occupational class, NS-SEC 1 or higher managerial and professional occupations, which have long been subject to policy concerns about social exclusivity and elitism (Milburn, 2009, 2012, 2013). Indeed these findings clearly puncture romantic notions of the CCIs as an exemplar of merit and accessibility and instead point towards a sector dominated by the children of managers and professionals.
Are the creative industries meritocratic? An analysis of the 2014 British Labour Force Survey
Published online:
12 April 2016Table 2. Origins in creative industries compared with population and NS-SEC 1 and 2.
One of the advantages of the large-sample LFS is that it allows for an unusually detailed investigation of how the distributions of class origins vary across different CCI sectors. Table 3 suggests that the CCIs are in no way a coherent formation in terms of their social composition. Some sectors, such as publishing, advertising, and music and performing and visual art, have a particularly high concentration of those from professional and managerial backgrounds (NS-SEC 1 and 2) whereas the distribution of the origins of those working in craft, by contrast, is much closer to what is found in the general population.
Are the creative industries meritocratic? An analysis of the 2014 British Labour Force Survey
Published online:
12 April 2016Table 3. Origins for each sector and occupation.
Class, gender, and pay
While Table 3 describes important variations in occupational “openness” across different CCIs, another pressing question – in terms of meritocracy at least – is whether earnings variation exists for employees within the CCIs according to gender and class origin.2 We use a series of multivariate linear regression models of earnings in the CCIs to answer this question. Table 4 reports the results of regressions: in the first column is a model with only measures of class origin and gender. The next model adds controls for ethnicity, age, country of birth, and hours worked (as well as a control for the wave in which the respondent answered income questions, not shown). The third column includes measures of a host of other factors known to affect earnings: working in London, education, firm size, public vs. private sector, job tenure, training, and specific occupation within each creative sector (not shown in the table). Many of these items (such as educational credentials and whether or not one works in London) are associated with class origins (see appendix table A2 and Friedman et al., 2016).3
Are the creative industries meritocratic? An analysis of the 2014 British Labour Force Survey
Published online:
12 April 2016Table 4. Models of earnings in the CCIs.
The hidden barriers, or “glass ceiling”, preventing women from getting to the top of the CCIs are well documented (Conor, Gill, & Taylor, 2015; Gill, 2014; Scharff, 2015; Skillset, 2010). Table 4 shows that this glass ceiling or gender pay gap is emphatically confirmed in the LFS data: female employees have average earnings of £239/week or over £12,000/year less than men (with similar class backgrounds) in the CCIs as a whole. Some of this pay gap is accounted for by differences between men and women in the CCIs. Women in the CCIs are on average younger than men, more likely to work in the public sector and in other less-well-paid occupations within the CCIs. Thus, women have predicted earnings of £130/week less than men who are otherwise similar on the measures in the base model, and £112/week less than men net of all the controls in the full model. Nonetheless, net of all these controls there is still a substantial and statistically significant pay gap, with women employed in CCIs earning about £5800 less per year than otherwise similar men.
In keeping with the rest of the analysis presented here, however, the picture is more mixed when drilling down into individual sector data. Table 5 reports the gender pay gap in each of the nine sectors of the CCIs in three ways: the pay for men and the difference for women (Column 1) without any controls; the difference between men and women after demographic and hours-worked controls only (Column 2, the same model as the “base model” in Column 2 of Table 4); and the full model (Column 3, again the same model as Column 3 in Table 4). We find that statistically significant gender pay gaps persist in Architecture, Crafts, Film TV & Radio, and IT, with estimates ranging from £97 to £288/week, or from about £5000/year in IT to nearly £15,000/year in Film and other media. The pay differences between men and women in Advertising, Design, Publishing, and Museums & Galleries, on the other hand, are not statistically significant in the full model.4 In sum, though, the new LFS figures underline the striking scale of disadvantage faced by women employed within the CCIs. While a gender pay gap does not represent a new finding, our analysis gives the most granular understanding to-date of how gender inequality plays out in different CCI occupations.
Are the creative industries meritocratic? An analysis of the 2014 British Labour Force Survey
Published online:
12 April 2016Table 5. Gender pay gaps.
Work examining earnings inequality by class origin within the CCIs is less developed. Yet elsewhere in British sociology, Laurison & Friedman (2016) have demonstrated that even when those from working-class backgrounds do successfully enter high-status occupations they have, on average, considerably lower incomes. At present there is little understanding of whether this “class ceiling” extends to the CCIs. The new LFS data we present here, then, with its detailed and accurate measures of parental occupational class and employees’ individual earnings, represents a unique opportunity to address this gap. Returning to Table 4, this suggests that there is a class-origin pay gap within Britain’s CCIs. The first and second models return statistically significant and substantively meaningful differences in earnings between employees with parents in NS-SEC 1 occupations (the reference category) and those from NS-SEC 2 or NS-SEC 6–8 backgrounds: workers from working-class origins have earnings on average £157/week or over £8100 less per year than demographically similar people (working the same number of hours) from privileged backgrounds. However, these differences are much smaller, and statistically insignificant, in the full model; this suggests that earnings differences by origin are accounted for by differences in the educational levels, particular work contexts and occupations of those from working-class origins. That differences in earnings may be accounted for by education levels and working context should, in itself, be a major cause for concern given what recent sociological research has suggested on the links between class and education, class and occupation, and class and geography (Savage et al., 2015).
Finally, in Table 6 we turn to the evidence of “class ceilings” in individual sectors of the CCIs. Here, we show the differences between employees from privileged NS-SEC 1 origins and everyone else; that is, employees whose parents were in any other NS-SEC category below NS-SEC 1. There are statistically significant differences in pay, net of all controls, for those from backgrounds outside NS-SEC 1, in Film, IT, and Publishing, ranging from £117/week to £444/week or about £23,000/year. Conversely, there is also a class origin bonus of £189/week in Advertising.
Are the creative industries meritocratic? An analysis of the 2014 British Labour Force Survey
Published online:
12 April 2016Table 6. Class-origin penalties.
All of these estimates of class and gender differences in earnings, especially in individual sectors with small numbers of respondents, are necessarily approximate, and change somewhat depending on the particular covariates in each model. Further, none of these regression models can identify the causes of these discrepancies. They are, however, clear indications that there are both gender and class-origin income inequalities facing employees in many of the CCIs, and that these are not accounted for by measurable differences between women and men, or between people from different class origins.
Conclusion: rethinking CCIs
DCMS’ (2015a, 2015b) most recent economic estimates suggest CCIs are a well performing area of the economy. This makes them highly attractive to policy-makers looking for a vision of the future for a British economy increasingly dependent on service sector occupations (Engelen et al., 2011). Indeed, the findings from this paper support a picture of CCIs as well remunerated and even, in the case of advertising, offering better pay to those employees from non-elite class origins, as compared to those from more elite starting points. However, the main thrust of our argument is that important questions remain about how far the economic success of the CCIs rests on relatively socially homogenous foundations. Moreover, our analysis demonstrates very clearly that even when women and those from working-class backgrounds are able to make it as employees in this sector they face significantly lower average wages as compared to the middle class and compared to men.
The uneven distribution of diverse social groups working within CCIs, alongside the disparities in rewards, suggests two things. First that inequalities in the CCI labour force are of an intersectional nature (an issue addressed at length by Oakley & O’Brien, 2016). This begins, as Table 4 shows, with educational inequalities and is compounded by the uneven geography of access to creative work, specifically whether employees are working in London. Altogether, the results of our analyses indicate a complex configuration of interactions between class, ethnicity, and gender, across a range of CCIs that have very different occupational cultures. More work, of the kind highlighted by Conor et al. (2015) and Hesmondhalgh and Saha (2013), is needed to fully detail both the operation of these intersections and how they vary over time and across different jurisdictions (see e.g. Koppman’s work on American CCIs, 2015).
This latter point gestures towards the second implication of our findings in this paper. There is a clear question raised as to the coherence of aggregating CCI occupations into a single sector of the economy. Here our analyses speak directly to a longstanding concern within academic literatures on CCIs that the occupational groups included in their definition are simply too distinct from each other to represent a coherent sector of the economy. This conclusion draws attention to the need for a decomposition of CCIs and attentiveness to the diversity within and between the individual parts. While this has traditionally been approached on an occupation by occupation basis, understanding the components of the CCIs relationally has become central to recent calls from cultural studies scholars to better understand the future role of CCIs in economy and society (McRobbie, 2015).
Research on inequalities in recruitment with a particular focus on pay gaps (particularly those associated with social class origin) has, until now, not featured in studies of CCIs. While DCMS has published estimates focusing on employment (2015b), in the absence of any concerted analysis of the social origins of the CCI workforce, the central thrust of the government’s work has been on the “good news” of CCIs’ economic contribution and their seemingly meritocratic profile.
In contrast, our analyses show clear and often striking inequalities across and between the CCIs. Some of these concern longstanding issues discussed by policy and by the media, such as the under-representation of women. However, introducing class origin into these debates – as we have done here – raises a number of new and important questions about the particular nature and consequences of inequality within the CCIs. For example, given the dominance of the children of professionals and managers in publishing, what are the implications for English literary culture? What might be the consequences for the way we think about cultural value (O’Brien & Lockley, 2015), if the sector is so unevenly accessible for those from different backgrounds?
Moreover, the “good” economic news about the CCIs is substantively driven, as Campbell (2014) has noted, by two sectors: IT and Advertising. Our analyses in Table 1 and Figure 1 above reiterate the longstanding point that IT, Software, and Computer Services is by far the biggest employer in the sector, followed by advertising. These are also the best-remunerated occupational groups. The fact that they represent the best-paid cultural professionals, and together make up the 53% of the total CCI workforce, clearly shows how their inclusion skews any understanding of the economic contribution and potential of the CCI sector as a whole.
This skewing of the CCIs economic contribution is echoed in our analysis of their in internal inequalities. As we noted with regard to Table 3, CCIs do not exhibit uniform patterns with reference to the social origins of their workers. This internal differentiation not only points to important differences in the relative openness of different CCI occupations but also shows how the DCMS aggregation of the CCI sector hides significant inter-occupational class inequalities.
What our research here suggests is that, following Miles (2016), there are very different occupational cultures within DCMS’ CCIs, whether in terms of the social origins, genders, or ethnicities of the workers, or in terms of their remuneration. This indicates that sector analysts and policy-makers need to re-open the definitional debates that organisations such as NESTA, with the idea of “creative intensity”, hoped to settle. This is not for reasons of nitpicking or academic quibbling but because, as the LFS shows, we are talking about occupations that are profoundly different from one another. For example pay rates, ethnic diversity, and class origins are vastly different in IT as compared to publishing, while educational attainment levels in craft are clearly different to those in architecture. Which policy and practice frameworks offer the most appropriate understanding of CCIs therefore remains open to question, particularly when thinking about regulation of employment practices, such as internships or low or no pay forms of work, which are the basis for entry into many cultural occupations. This question is thrown into particularly sharp relief when considering the ability, or not, of CCIs to deliver on the promise of a meritocratic, socially mobile, and well remunerated new economy, given the inequalities so clearly displayed by the current labour force.
| % male | % white | % w/Uni degree or higher | Weighted per cent of CCIs | n | n with income | |
|---|---|---|---|---|---|---|
| IT, software, and computer services | 85.2 | 85.5 | 68.4 | 29.80 | 478 | 348 |
| Advertising and marketing | 51.7 | 92.6 | 65.5 | 23.20 | 372 | 244 |
| Music, performing and visual arts | 55.1 | 92.9 | 60.4 | 8.90 | 147 | 14 |
| Design: product, graphic, and fashion design | 53.7 | 96.1 | 55.3 | 8.10 | 136 | 69 |
| Publishing | 45.6 | 93.1 | 71.0 | 7.80 | 133 | 55 |
| Film, TV, video, radio, and photography | 75.9 | 93.2 | 59.6 | 7.30 | 117 | 37 |
| Crafts | 76.6 | 95.0 | 20.6 | 6.10 | 106 | 41 |
| Architecture | 70.1 | 91.7 | 87.2 | 6.40 | 105 | 74 |
| Museums, galleries, and libraries | 37.3 | 97.5 | 76.3 | 2.40 | 43 | 36 |
| Total in CCIs | 65.8 | 91.1 | 64.0 | 100.00 | 1637 | 918 |
| Total in NS-SEC 1 | 65.7 | 89.9 | 66.3 | |||
| Total in NS-SEC 2 | 44.8 | 91.7 | 48.5 | |||
| Total in population | 49.1 | 89.8 | 29.7 |
Notes: Weighted percentages based on recommended survey weighting. All respondents reporting an occupation assigned to a CCI sector, aged 23–69 and not in full-time education.
| NS-SEC1 Origins | NS-SEC2 Origins | NS-SEC3–5 Origins | NS-SEC6–8 Origins | |
|---|---|---|---|---|
| Creative industries as a whole | 26.1% | 23.9% | 32.0% | 18.0% |
| NS-SEC 1 | 26.4% | 20.6% | 33.5% | 19.5% |
| NS-SEC 2 | 18.3% | 20.2% | 35.9% | 25.7% |
| Population as a whole | 14.1% | 15.0% | 36.2% | 34.7% |
Notes: Weighted percentages based on recommended survey weighting. All respondents reporting an occupation, aged 23–69 and not in full-time education. N = 1637 for CCIs, 5491 for NS-SEC 1, 9614 for NS-SEC 2, and 45,356 for population as a whole.
| Higher Prof & Mgrs | Lower Prof & Mgrs | Intermediate Occs | Routine and semi-routine | n | |
|---|---|---|---|---|---|
| Publishing | 43.2% | 17.7% | 27.1% | 11.9% | 133 |
| Authors, writers, and translators | 47% | 15% | 28% | 10% | 72 |
| Journalists, newspaper, and periodical editors | 38% | 20% | 26% | 15% | 61 |
| Advertising and marketing | 30.8% | 24.0% | 26.3% | 19.0% | 372 |
| Public relations professionals | 38% | 26% | 24% | 12% | 51 |
| Marketing and sales directors | 33% | 23% | 27% | 17% | 137 |
| Advertising accounts mngrs and creative directors | 29% | 22% | 28% | 20% | 24 |
| Advertising and public relations directors | 28% | 22% | 29% | 21% | 24 |
| Marketing associate professionals | 27% | 24% | 25% | 23% | 136 |
| Music, performing and visual art | 28.3% | 25.0% | 32.9% | 13.8% | 147 |
| Musicians | 38% | 21% | 30% | 10% | 54 |
| Dancers and choreographers | 24% | 24% | 41% | 11% | 9 |
| Actors, entertainers, and presenters | 22% | 29% | 29% | 21% | 36 |
| Artists | 21% | 27% | 38% | 14% | 48 |
| Design: product, graphic, and fashion design | 26.1% | 19.2% | 33.6% | 21.2% | 136 |
| Graphic designers | 32% | 15% | 32% | 21% | 69 |
| Product, clothing, and related designers | 20% | 23% | 35% | 22% | 67 |
| Architecture | 24.3% | 24.0% | 38.2% | 13.5% | 105 |
| Architects | 27% | 20% | 44% | 10% | 53 |
| Chartered architectural technologists | 27% | 0% | 73% | 0% | 3 |
| Town planning officers | 25% | 29% | 24% | 22% | 26 |
| Architectural and town planning technicians | 18% | 31% | 37% | 14% | 23 |
| IT, software, and computer services | 22.5% | 25.4% | 32.6% | 19.6% | 478 |
| Web design and development professionals | 32% | 28% | 29% | 11% | 57 |
| Programmers and software development professionals | 21% | 28% | 32% | 18% | 245 |
| IT and telecommunications directors | 21% | 21% | 36% | 21% | 83 |
| IT business analysts, architects and systems designers | 21% | 21% | 32% | 27% | 93 |
| Museums, galleries, and libraries | 27.8% | 24.5% | 22.2% | 25.6% | 43 |
| Archivists and curators | 41% | 27% | 15% | 17% | 13 |
| Librarians | 21% | 23% | 25% | 30% | 30 |
| Film, TV, video, radio, and photography | 17.2% | 35.2% | 38.4% | 9.1% | 117 |
| Arts officers, producers, and directors | 20% | 37% | 39% | 3% | 53 |
| Phtgrphrs, AV and broadcasting equipment oprtrs | 15% | 33% | 38% | 14% | 64 |
| Crafts | 12.9% | 14.8% | 43.1% | 29.2% | 106 |
| Smiths and forge workers | 39% | 0% | 31% | 30% | 3 |
| Glass and ceramics makers, decorators, and finishers | 20% | 14% | 38% | 28% | 17 |
| Furniture makers and other craft woodworkers | 13% | 20% | 42% | 25% | 38 |
| Other skilled trades n.e.c. | 9% | 12% | 48% | 31% | 42 |
| Weavers and knitters | 0% | 14% | 34% | 52% | 6 |
Note: The low N for the occupations in italic font reflects the difficulty of drawing sector wide conclusions. The rows in bold represent the overall CCI sectors
| 1 | 2 | 3 | |
|---|---|---|---|
| Variable | No controls | Basic controls | All controls |
| Female (vs. Male) | −239*** | −130*** | −112** |
| Social Origins (vs. NS-SEC 1 origins) | |||
| NS-SEC 2 parents | −110** | −80* | −40 |
| NS-SEC 3–5 parents | −29 | −67 | 43 |
| NS-SEC 6–8 parents | −132** | −157*** | −21 |
| BAME (vs. White) | −44 | −27 | |
| Age | 86*** | 48*** | |
| Age Squared | −1*** | −0*** | |
| Birth Country | |||
| Outside UK | 96 | 20 | |
| Northern Ireland | 11 | 49 | |
| Scotland | −73 | −36 | |
| Wales | 77 | 7 | |
| Paid Hours worked | 18*** | 12*** | |
| Degree (vs. University Degree) | |||
| Less than Uni Degree | −113** | ||
| Postgrad | 17 | ||
| Work in London | 139*** | ||
| Public Sector worker | −195*** | ||
| Firm Size (vs. less than 25 employees) | |||
| 25 to 49 employees | 103** | ||
| 50 to 499 employees | 81** | ||
| 500 or more employees | 266*** | ||
| NS-SEC Category (vs. Higher Managers & Administrators, 1.1) | |||
| Higher Professionals (1.2) | −337* | ||
| Lower Managers & Professionals (2) | −475*** | ||
| Everybody else (3–8) | −562*** | ||
| Years at Current Job | 0 | ||
| Participated in Job-Related Training last 3 months | 38 | ||
| Constant | −1641*** | −35 | |
| N | 877 | 862 |
Notes: The second column shows the results of a regression with only demographic variables and number of hours worked (plus dummy variables for which the quarter in which the respondent reported their income and occupation, not shown). The second column includes a full range of controls, including dummies for quarters and individual occupations within the CCIs (not shown).
*p < .10.
**p < .05.
***p < .01.
| 1 | 2 | 3 | |||
|---|---|---|---|---|---|
| Men | Women, no controls | Women, base controls | Women, net of controls | n, full model | |
| Advertising and marketing | 1163 | −458*** | −261** | −105 | 232 |
| Architecture | 732 | −137** | −51 | −151** | 72 |
| Crafts | 439 | 139 | 6 | −272** | 39 |
| Design: product, graphic, and fashion | 593 | −128* | −85 | −109 | 67 |
| Film, TV, video, radio, and photography | 773 | −83 | −4 | −288** | 35 |
| IT, software, and computer services | 906 | −211*** | −114** | −97* | 336 |
| Publishing | 770 | −74 | 63 | −47 | 53 |
| Museums, galleries, and libraries | 402 | 15 | −42 | −1 | 34 |
| Music, performing and visual arts | 450 | −194 | −405*** | 9879*** | 14 |
Note: The first column gives the average income for men, no controls; the next column gives the difference between the average earnings for men and those for women without any controls. The third and fourth columns give the coefficients for women in each sector from models with the same covariates as columns 2 and 3 of Table 4, respectively. Column 4 thus gives the estimated gender penalty net of controls for age, class origin, working in London, ethnicity, education, hours worked, firm size, public vs. private sector, job tenure, training, and specific occupation within each creative industry. The low N for the occupations in italic font reflects the difficulty of drawing sector wide conclusions.
*p < .10.
**p < .05.
***p < .01.
| 1 | 2 | 3 | |||
|---|---|---|---|---|---|
| No controls | Base controls | Full controls | |||
| NS-SEC 1 origins | NS-SEC 2–8 Origins | NS-SEC 2–8 Origins | NS-SEC 2–8 Origins | n | |
| Advertising and marketing | 871 | 103 | 112 | 189* | 232 |
| Architecture | 709 | −32 | −63 | −8 | 72 |
| Crafts | 541 | −94 | −65 | 78 | 39 |
| Design: product, graphic, and fashion | 534 | 9 | −113** | −27 | 67 |
| Film, TV, video, radio, and photography | 1067 | −401** | −459** | −444*** | 35 |
| IT, software, and computer services | 1011 | −180** | −198** | −117* | 336 |
| Publishing | 805 | −162 | −269** | −191* | 53 |
| Museums, galleries, and libraries | 367 | 61 | −45 | 3 | 34 |
| Music, performing & visual arts | 329 | 34 | −131 | 14 | |
Note: The first column gives the average income for people from NS-SEC 1 origins, no controls; the next column gives the difference between the average earnings for privileged-origin people and those with parents in any other NS-SEC category, without any controls. The third and fourth columns give the coefficients for NS-SEC 2–8-origin people in each sector from models with the same covariates as columns 2 and 3 of Table 4, respectively. Column 4 thus gives the estimated class-origin penalty net of controls for age, class origin, working in London, ethnicity, education, hours worked, firm size, public vs. private sector, job tenure, training, and specific occupation within each creative industry. The low N for the occupations in italic font reflects the difficulty of drawing sector wide conclusions.
*p < .10.
**p < .05.
***p < .01.
O_Brien_05c_ONLINE_appendix.docx
Download MS Word (36 KB)Notes
1. Although it is standard in mobility table analyses to focus on those who are 35 or older and have likely landed in a stable career, we include the widest reasonable age range because we are interested in the composition of the creative and cultural industries’ workforce.
2. It would be ideal to also examine differences between whites and BAME people, however, the number of non-white respondents in most sectors is too small for meaningful analyses.
3. The rationale for including each of these measures is discussed fully in Laurison & Friedman (2016); some of these variables are properly thought of as controls, such as age, and others (such as education, and, as our other research has shown, working in London) mediate the relationship between class origin and earnings.
4. We also report the results of models for Music and Arts, but the very small number of respondents with income information in this group makes these results unreliable.
