Acknowledgements

The authors would like to thank Douglas Cameron and Tom Knight of the Department for Culture, Media & Sport for their assistance in producing the data tables and figures in this report, and Mark Spilsbury and Derek Bosworth for their comments on an earlier version.

Nesta is an innovation charity with a mission to help people and organisations bring great ideas to life.

We are dedicated to supporting ideas that can help improve all our lives, with activities ranging from early-stage investment to in-depth research and practical programmes.

Nesta is a registered charity in England and Wales with company number 7706036 and charity number 1144091. Registered as a charity in Scotland number SC042833. Registered office: 1 Plough Place, London, EC4A 1DE.

www.nesta.org.uk ©Nesta 2015

EXECUTIVE SUMMARY

The UK's creative and high-tech economies are major employers. Between them, and without double counting, they account for 4.76 million jobs, or around 16 per cent of the UK workforce (average of 2011-2013). Within this, 2.52 million jobs are in the creative economy, 3.11 million are in the high-tech economy, and 0.87 million are in both.

Between 2011 and 2013, employment in the UK's creative economy (jobs in the creative industries plus creative jobs elsewhere) grew by 4.3 per cent per annum (p.a.) on average, over three times faster than the workforce as a whole (1.2 per cent p.a.). This is also faster than employment growth in the high-tech economy – that is, jobs in the high-tech industries plus Science, Technology, Engineering and Mathematics (STEM) jobs outside the high-tech industries – which grew at 2.1 per cent p.a.

Similarly, employment in the UK's creative industries grew at 5.0 per cent p.a., much faster than the 1.5 per cent p.a. growth of employment in the high-tech industries (which was dragged down by particularly slow growth in non-STEM occupations).

Echoing Nesta's previous research findings using business registry data, we detect strong tendencies for creative and high-tech employment to co-locate. The creative economy is, however, less equally distributed across the UK, with London being more prominent than it is in the UK's high-tech economy. There is some evidence though of a slight catch up since 2011 as London's creative economy workforce has grown more slowly than most parts of the country.

How big are the UK's creative and high-tech economies?

There were 2.52 million jobs in the UK's creative economy (average 2011-2013), representing 8.3 per cent of the workforce. This consisted of 1.65 million jobs in the creative industries (866,000 of which were in creative occupations and 782,000 of which were in other roles) and 876,000 jobs in creative occupations outside of the creative industries (sometimes called 'embedded creatives'). That is, there are slightly more creative jobs outside the UK's creative industries than within.¹

There were 3.11 million jobs in the high-tech economy (average 2011-2013), around 10.3 per cent of the workforce. This consisted of 2.33 million jobs in high-tech industries (805,000 of which were in STEM occupations and approximately 1.53 million in other roles) and 782,000 STEM jobs outside the high-tech industries. The extent to which STEM jobs are embedded in the wider UK economy is therefore similar in proportional terms to creative jobs.

0.87 million jobs fell within both the creative and high-tech economies (average 2011-2013, 2.8 per cent of the workforce); 0.46 million of these were at the same time both creative and STEM occupations. 0.54 million people worked in industries that we classify as both creative and high-tech. That part of the UK workforce where the creative and high-tech economies meet has been particularly dynamic, growing faster than the workforce as a whole over the 2011-2013 period at 8.0 p.a. on average (with dual creative/high-tech occupations growing at 5.7 per cent p.a. and dual creative/high-tech industry employment growing especially rapidly at 9.6 per cent p.a.). Policymakers should therefore pay particular attention to this segment of the UK's economy.

What is the geographic spread of creative and high-tech employment and how is it changing?

Regions with the highest concentrations of creative economy employment tend also to be regions with high concentrations of high-tech economy employment (concentrations being defined in terms of the level of creative (high-tech) economy employment as a proportion of the area's workforce).

The creative economy is, however, more unequally distributed across the country than is the high-tech economy. In particular, it is even more highly concentrated in London (where it accounts for 15.5 per cent of the workforce) compared with the high-tech economy which employs 10.6 per cent of the workforce in London.

At a sub-regional level the creative and high-tech economies are particularly concentrated in the counties to the north, south and west of London (Berkshire, Oxfordshire, Buckinghamshire, Surrey, Hertfordshire, Cambridgeshire) in Outer London (West and North West) and in Milton Keynes, Edinburgh and Bristol.²

There are pockets of concentrated high-tech economy employment without correspondingly high concentrations of creative employment in areas including Aberdeen City & Aberdeenshire, Cheshire, Derby, West Cumbria, Swindon, Halton and Warrington, Bedfordshire, Hampshire and Warwickshire.

Sub-regional concentrations of the creative economy where there is no corresponding concentration of high-tech can be found in Inner London (East and West), and in Outer London (South). Brighton and Hove is also notable for its concentration of creative economy employment. This is not deny the existence of high-tech activity in Inner London, but as our definition is broader encompassing a range of different technological activities it does not pick these out in isolation.

Over the period 2011 to 2013 London's creative economy grew on average by 2.9 per cent p.a., just under three-quarters of the rate of the UK's creative economy as a whole (4.3 per cent p.a.). This is all the more striking given that London's overall workforce grew almost twice as fast as the UK's over this period. With the exception of Scotland, the creative economy grew more rapidly in all areas outside London, particularly the East of England (9.3 per cent p.a.), the West Midlands (8.2 per cent p.a.) and the North East (5.6 per cent p.a.).

Employment in London's high-tech economy grew more quickly than did its creative economy, at 4.5 per cent p.a., more than double the 2.1 per cent p.a. growth achieved in the UK as a whole. With the exception of the South West of England and Northern Ireland, all UK regions saw expansion in their high-tech economies.

1. INTRODUCTION

In recent years, analysts have variously employed terms such as the knowledge economy,³ information economy,⁴ creative economy,⁵ and high-tech economy⁶ to characterise a trend in advanced developed countries towards economic activities where conceptions of technical know-how, expertise and creativity are considered critical. The two concepts of creative and high-tech industries have had particular longevity – in the UK and beyond – though definitional ambiguities have prevented an understanding of how they compare. This report corrects these ambiguities, clarifies the relationship between creative and high-tech industries and compares their size and recent performance.

In policy circles, it is widely believed that both these industries enjoy faster productivity growth than other sectors and are therefore good for a country's international competiveness and economic growth. They are also viewed as engines of job growth as they employ highly-skilled workers who are less vulnerable to substitution from new technologies.⁷ Last, but not least, both sectors are seen as supporting growth in the wider economy, insofar as they create knowledge spillovers, such as those embodied in mobile labour, which drive innovation and productivity growth in other industries. These spillovers, it is argued, constitute a market failure and justify policy intervention. As such, policymakers in an increasing number of countries – developing as well as developed are prioritising their creative and high-tech industries for policy support (UNESCO, 2013, Edwards and Lawrence, 2010).

The importance of skilled labour and human capital is reflected in an emphasis on talent in the definitions used to understand these industries, as illustrated, for example, by the Department for Culture, Media & Sport (DCMS) definition of the creative industries, and the role assigned to STEM knowledge or skills in many accounts of the high-tech industries. Thus, the DCMS's 1998 Mapping Document, the first national attempt to define the UK's creative industries, defined them as:⁸

"those industries which have their origin in individual creativity, skill and talent and which have a potential for wealth and job creation through the generation and exploitation of intellectual property

and a US Congressional Office of Technology Assessment document describes high-tech firms as those:⁹

engaged in the design, development, and introduction of new products and/or innovative manufacturing processes through the systematic application of scientific and technical knowledge

In our 2013 study, A Dynamic Mapping of the UK's Creative Industries,¹⁰ we noted that concepts like the 'creative industries' and 'creative economy' – indeed 'creativity' itself – although widely used by policymakers, lacked sufficiently clear and rigorous definitions. The report aimed to address this weakness by introducing clear criteria for which occupations should or should not be considered 'creative' for measurement purposes. The report also introduced a methodology for determining which industries should be classified as 'creative' based on their creative intensity (Freeman 2004, 2008a, Bakhshi et al., 2013:12), defined as the proportion of an industry's workforce in creative occupations. It found that creative intensity distinguishes 'creative industries' from other industries which also employ creative talent: although large numbers of individuals are employed in creative occupations in many industries, only a small number of these – the 'creative industries' – have exceptionally high levels of creative intensity.

Concepts like 'high-tech industry', and 'STEM' occupation, it turns out, are also plagued by definitional ambiguities and uncertainties. The wide range of classifications adopted by different studies – even within countries – has given rise to significant differences in estimates of the size, growth and geography of the high-tech industries and also of the demand for, and use made of, STEM occupations and skills in the wider economy.

The idea of using STEM workforce intensity (the proportion of an industry's workforce in STEM occupations) to identify high-tech industries has been independently applied by a number of US researchers. For example, Hecker (2005:58) defines high-tech industries in the US as follows:

An industry is considered high tech if employment in technology-oriented occupations accounted for a proportion of that industry's total employment that was at least twice the 4.9 per cent average for all industries.¹¹

In this report, we ground our definition of STEM occupations, the high-tech industries and the wider high-tech economy, in a statistical analysis of STEM intensity which mirrors our use of creative intensity, but which differs from it in significant ways (and in particular, as discussed below, reflecting the fact that STEM intensity turns out to be a weaker discriminator between industries than creative intensity).

In summary, in this report we make the following contributions:

1. We review the classifications of the creative and high-tech economies in the UK context (Section 2).
2. We update the baseline estimates of the UK's creative industries and creative economy in Dynamic Mapping for the period 2011-2013 and analyse their growth rates and geographies (Section 5).¹²
3. We estimate the number of individuals employed in STEM occupations in the UK and analyse how STEM intensity is distributed across industries. We find that the tendency for a small set of industries to have particularly high proportions of STEM workers in their workforce is less pronounced than is the case for creative occupations in the creative economy (Section 6).
4. We combine the STEM intensity analysis with Eurostat's classifications of high-tech industries, based on their level of R&D spend (for manufacturing) and knowledge intensity, measured by the qualifications of their workforce¹³ for (services), to generate a classification of high-tech industries, and corresponding estimates of the size, growth and regional geography of high-tech industries and the high-tech economy (Section 6).
5. Using our classifications for creative and high-tech industries and economies we compare the make-up of these two parts of the UK economy and describe how they relate to one another in accounting terms (Section 7).
6. We analyse the sub-regional geography of the UK's creative and high-tech economies and compare it with how other sectors are distributed (Section 8).
7. We conclude by summarising the research findings and discussing the implications for measurement and future research (Section 9).

The next section provides some of the UK context on debates surrounding the definition of the creative and high-tech economies.

2 POLICY CONTEXT

2.1 DEFINITIONS OF THE UK'S CREATIVE ECONOMY

The 1998 and 2001 DCMS Mapping Documents¹⁴ proposed that 13 sub-sectors made up the creative industries. These industries appeared to form a reasonably coherent group based on the generation of commercial value by creative talent. The 13 sub-sectors however, contained a notable split between household-facing sub-sectors like music, film and performing arts and business-facing sectors like advertising, design and software, leading to accusations of over-inclusion and boosterism.¹⁵

There were also many caveats and limitations in the DCMS Mapping Documents concerning gaps, a lack of comparability across sub-sectors, and the reliance on highly aggregated source data. The annual DCMS Creative Industries Economic Estimates, first published in 2002, attempted to address some of these challenges, at least for those sub-sectors identifiable using the official Standard Industrial Classification (SIC) codes, by producing each year the same headline statistics on Gross Value Added (GVA), employment, net exports and the number of creative businesses, based on consistent data sources for each sub-sector. These headline statistics showed the UK's creative industries were growing at twice the rate of other sectors, helping to raise the profile of the sector.¹⁶

A great many challenges remained in the DCMS statistics as a result of the limitations of the SIC codes, which are set in conjunction with the UN's industrial classification system for the purposes of international consistency and which are reviewed only at roughly ten-year intervals.

Partly because of this – but also due to differing data needs – sector bodies such as UK Music, the British Fashion Council and the Crafts Council started producing their own sector-specific economic statistics, along with equivalent bodies across the devolved nations and the English regions. Inconsistent treatments of sectoral boundaries (what should be included and what should not) led to a plethora of non-comparable estimates. The result was an ever-increasing landscape of sectoral, national and regional statistics that purported to measure similar things but were not strictly comparable.¹⁷

The position was exacerbated further in December 2011 when the DCMS removed two software-related occupations and industries from its classifications, cutting £25.9 billion from the amount creative industries were estimated to contribute to UK GVA. This decision showed a lack of appreciation for the interconnected relationship between software and creative content. But it also exposed the fact that the DCMS statistics had always been based on a pragmatic selection of occupational and industrial codes, and not on a systematic methodology.

Dynamic Mapping proposed an explicit methodology for determining which occupations and industries should be classified as creative for measurement purposes. It assessed the different characteristics which make an occupation creative, and then used the workforce intensity of employment in creative occupations to identify creative industries.¹⁸ By showing that a defining characteristic of creative industries was their especially intensive use of creative talent, it provided a statistical basis for treating the different creative industries as a coherent group. Following a public consultation exercise, in January 2014 the DCMS adopted this methodology for its official creative industries economic estimates.¹⁹

2.2 DEFINITIONS OF THE UK'S HIGH-TECH ECONOMY

Policymakers in the UK have long held the view that the UK's future growth prospects depend on it having a strong high-tech economy. The 2011 HM Treasury/BIS Plan for Growth states that:

“Britain has to earn its way in the modern world. We have to become much more productive so we can be a leading high tech, highly skilled economy. We must build a new model of economic growth where instead of borrowing from the rest of the world, we invest and we save and we export. Our economy must become more balanced.²⁰

This quote brings out a number of aspects of UK policymakers' interest in high-tech: the idea that it is a future area of opportunity, the association of high-tech with high skills, and the view that the UK economy needs to be more 'balanced'.²¹

However, the rhetoric of 'high-tech' has outpaced improvement in its measurement. Its scope is rarely defined, statistics on its economic importance are lacking and, in contrast to the creative economy, there is no 'official' classification. A review of the use of the term in the UK public policy literature reveals that it has been variously thought of as:

• A set of technologies.
• A group of firms that invest in high levels of R&D activity.
• A key employment destination for Science Technology Engineering and Mathematics (STEM) talent.
• A distinct group of industries.

We discuss these in turn:

A set of technologies

In 2012 David Willetts, the Minister of State of Science and Universities, in a speech on the UK's high-tech future²² referred to the Government Office for Science's 2010 report identifying future technologies and innovations likely to be important to the UK in the 2020s. These being classified into the following broad categories:²³

1. Biotechnology and Pharmaceuticals.
2. Materials and Nanotechnology.
3. Digital and Networks.
4. Energy and low-carbon technologies.

In a subsequent publication, Willetts outlined 'eight great technologies' where the government was working to ensure that they were properly supported, these being:

1. Robotics and autonomous systems
2. Satellites
3. Advanced materials
4. Regenerative medicine
5. Energy storage
6. Synthetic biology
7. Agri-science
8. Big data²⁴

A group of firms that invest in high levels of R&D activity

In 2010, a report by James Dyson set out a plan for making the UK the leading high-tech exporter in Europe.²⁵ This included a range of policy measures, among which were a number of support proposals for high-tech companies to encourage R&D investment. One of these was that the level of Enterprise Investment Scheme (EIS) tax relief should be raised for angel investors in high-tech companies. High-tech companies being defined on the basis of their level of R&D activities.

An employment destination for STEM talent

There is a long tradition in the UK of assessing whether the UK is producing enough people with STEM skills to meet its future workforce needs, resulting in successive reports by the Government on this subject: DTI (2006),²⁶ DFES (2006),²⁷ DIUS (2009)²⁸ and BIS (2011).²⁹

More recently, in 2011 and 2013 the UK Commission for Employment and Skills (UKCES) has undertaken reviews of the supply of and demand for high-level STEM skills in the UK.³⁰ The 2011 review was undertaken with the explicit aim of recognising that STEM skills support 'research, innovation and high-tech manufacturing'.³¹ It concluded that there was a broad match between the proportion of the workforce with high-level STEM skills and the proportion of jobs requiring such skills. However, it also found evidence of skills shortages in some sub-sectors, such as at graduate entry-level in engineering. The 2013 study, which contained projections out to 2020 under different economic growth scenarios, similarly concluded that although there were no STEM skills shortages in the UK workforce as a whole, certain regions and sectors regions suffered from skills gaps that held back growth.³²

The argument that specific high-tech industries are significant employers of STEM occupations has been made even more strongly in the US, as we discuss in Section 3 in more detail.

A distinct group of industries

An alternative approach has been to directly identify a set of industries as 'high-tech'. For example, KPMG's 2013 analysis of the location and changes in the UK's 'tech economy' defined technology industries by the following SIC codes, allowing their economic contribution to be estimated in the same way that DCMS does with the creative industries:

• Software publishing (SIC 582).
• Computer programming, consultancy and related activities (SIC 620).
• Data processing, hosting and related activities; web portals (SIC 631).
• Manufacture of computer, electronic and optical products (SIC 26).
• Manufacture of electrical equipment (SIC 27).

BIS's (2012) research paper on UK industrial strategy³³ assessed which high-tech sectors would in the future make the largest contribution to UK economic growth and employment. The report referred to high-tech in the context of 'high-medium tech manufacturing', though which industries codes were taken to make up this measure, and their current economic contribution to employment and value added, were not provided.³⁴

Having reviewed the UK context for defining the creative and high-tech economies, the next two sections outline the key features of the Dynamic Mapping methodology and the data sources for our analysis. The two subsequent sections present quantitative estimates of the size, growth and geography of the UK's creative and high-tech economies.

3 DYNAMIC MAPPING METHODOLOGY

The Dynamic Mapping methodology as originally applied to classify the creative economy consists of three stages.³⁵ First, a set of occupations are identified as creative. Second, the workforce intensity of these occupations is calculated for each industry in the economy. Third, based on the distribution of creative intensity across industries, a threshold intensity is identified, above which all industries are determined to be creative for measurement purposes and all those below are not.

Considering these stages in a little more detail:

1. Choice of occupation In the first stage, all occupations in the UK workforce as represented by Standard Occupational Classification (SOC) codes³⁶ describing the nature of a job are subjectively assessed on their creativity against five criteria.³⁷ A set of creative occupations is identified on this basis.
2. Calculation of intensity In the second stage, for every industry as defined by a set of Standard Industrial Classification (SICs) codes – the intensity i.e. the proportion of the industry's workforce employed in the selected occupations in the first stage is calculated. For example, SIC code 5912, the 'Motion Picture, video and television programme post-production activities' industry employed 12,000 people in 2012 (rounded to the nearest 1,000), of whom 8,000 were working in creative occupations. Accordingly, its creative intensity calculated from these rounded figures, is 8,000/12,000 = 67 per cent in that year.
3. Partitioning the industry according to an intensity threshold In the third stage, the distribution of creative occupation intensity is examined across all industries and a threshold intensity selected which is used to partition industries into two groups.

In Dynamic Mapping, the creative intensity threshold was selected by an intuitive equiprobabilistic criterion. That is, it was selected to be the intensity at which it was, on the best available estimate of the population parameters (based on the latest published DCMS estimates at the time of publication), equally likely that a creative industry would be wrongly rejected as not creative (a false negative) as would a non-creative industry be wrongly selected as being creative (a false positive). Appendix 4 contains further discussion of the issues raised in determining the threshold intensity in this report.

The set of occupations and industries generated by this analysis defines the employment in the 'economy' of interest. In the case of the creative economy, this consists of a combination of three types of employment:

1. Specialists: those working in creative occupations in creative industries.
2. Non-specialists: those working in a creative industry, but who are not themselves employed in a creative occupation.
3. Embedded: those working in creative occupations outside creative industries.

This has previously been described as the 'Creative Trident' (Higgs et al., 2005, Higgs et al., 2008).

In the present study, we ask if the same approach can be used to study the relation between the occupations and industries commonly considered part of the 'high-tech' economy. We begin by considering, in more depth, the appropriate group of occupations, starting from occupations commonly labelled as STEM occupations.

The connection between STEM intensity and high-tech industries has been studied quite extensively in the US. In 2004, the US Bureau of Labor Statistics (BLS) conducted an inter-agency seminar in which participants concluded that the Bureau should explore the whole range of factors contributing to the high-tech nature of an industry, including:

• A high proportion of scientists, engineers, and technicians (science, engineering, and technician occupation intensity).
• A high proportion of R&D employment (R&D employment intensity).
• Production of high-tech products, as specified on a Census Bureau list of advanced-technology products.
• Use of high-tech production methods, including heavy use of high-tech capital goods and services in the production process.

Following this, the BLS (Hecker, 2005) produced a classification of high-tech industries based on their intensity of high-tech employment.³⁸ High-tech industries were defined as those where the proportion of employment in the industry was at least twice the average (mean) of US industries as a whole. This classification was subsequently used by, among other studies, the Bay Area Council Economic Institute in their 2012 report on high-tech employment and wages in the United States.³⁹

Similarly, Echeverri-Carroll and Ayala, in their 2009 study of the effects of high-tech industry agglomerations on wages, classified as high-tech those manufacturing and service industries whose proportion of so-called 'technology orientated workers' was twice the national average or at least 6 per cent of their workforce.⁴⁰

While not using the language of 'high-tech', a much quoted study from the Brookings Institution (Rothwell 2013), uses an approach similar to Dynamic Mapping, by using the O*NET database (Occupational Information Network Data Collection Program) which contains information from extensive surveys of workers to classify occupations on the extent of their (in this case) STEM knowledge content.⁴¹

The report defines STEM occupations according to whether an occupation has a score at least 1.5 standard deviations from the mean for a given set of STEM-related criteria in O*NET. The resulting list of STEM occupations is then used to calculate a STEM intensity to identify industries with a particularly high use of STEM occupations, or, in other words, the high-tech industries.

In this report, we compute creative and STEM intensity for all UK industries at the 4-digit SIC level to identify, and therefore measure the economic contribution of, creative and high-tech industries, and the wider creative and high-tech economies they are part of.

Conceptually, it will help to fix ideas to establish a terminology that lets us discuss similarities and differences between these areas in a coherent way. If the relationship between STEM occupations and high-tech industries were the same as that between creative occupations and creative industries, then we could map ideas developed in the two areas of research as shown in Table 2.1.

TABLE 2.1 CONCEPTUAL SEGMENTATION OF THE CREATIVE AND HIGH-TECH ECONOMIES

The next section discusses the data sources we use in the analysis, and is followed by two sections that present estimates of employment in the creative and high-tech economies.

4 DATA SOURCES

The source for labour market data we use in this report is the Office for National Statistics' (ONS) Annual Population Survey (APS).⁴² The APS consists of the core Labour Force Survey (LFS) described below and a series of booster samples which are undertaken using a shorter questionnaire than the LFS. Currently, the quarterly Main LFS dataset contains around 120,000 individuals and the APS dataset contains around 340,000 individuals.⁴³ The results presented in this report are calculated using the APS weighted to gross up the figures to the population level.

The LFS is a quarterly household survey covering the UK; each year, the four quarters are combined to provide an annual survey. The LFS has the following characteristics:

• It covers both employee jobs and self-employment, which includes sole traders.
• It provides data on both the occupation of a person, and the industry that employs that person.
• It is repeated regularly – the surveys are conducted quarterly, but the data is amalgamated and joined with additional, so-called 'booster' surveys to produce an annual result.

Sections 5 and 6 discuss the creative and high-tech economies in turn. Each section covers the classifications we adopt, and the size, recent growth (since 2011) and the regional UK geography of the sector.

---

Footnotes

5. The Creative Economy

5.1 Classifying the Creative Economy

In January 2014, the DCMS adopted the Dynamic Mapping approach to identifying creative occupations and industries.⁴⁴

Uncertainties in the data and classifications meant that the DCMS considered it was important to review the inclusion or exclusion of some industries with creative intensities around the threshold level. The final classification of industries therefore also drew on feedback obtained from users of the statistics through a public consultation exercise (most obviously, SIC codes for museums and libraries were included on this basis). The consultation process, which led to the agreement of the final classification, was supported by cross-industry coordination overseen by the government's Creative Industries Council, including Nesta, the Sector Skills Councils (Creative Skillset and Creative and Cultural Skills), DCMS and a number of other industry bodies.

In this report, to be consistent with the resulting official creative industry statistics, we use the DCMS classifications for creative occupations and industries published in the January 2014 statistical release, even though there are some small differences with those in Dynamic Mapping. The corresponding 30 occupations and 31 industries are set out respectively in Tables 5.1.1 and 5.1.2 below, along with the creative intensities using 2013 APS data.⁴⁵

5.2 The Size of the Creative Economy

Based on these classifications, we estimate the overall size of the UK's creative economy, and its three main components: specialist, non-specialist, and embedded employment. Table 5.2.1 gives employment in the UK within these components for 2011 to 2013 and the average of these three years.

Note: The 2011-2013 average is given for comparison with later tables where greater disaggregation requires averaging over three years to avoid small sample biases.

The table shows that there were 2.6 million jobs in the UK's creative economy in 2013, and just over 2.5 million looking at the 2011-2013 period on average, of which 1.6 million were jobs in the creative industries.

The data can be arranged in a trident format with industries as columns and occupations as rows (Table 5.2.2).⁴⁶ This shows that around one-half of creative jobs were outside the creative industries in the wider creative economy.

It is common practice in presenting data on creative employment to divide these into sectors or segments.⁴⁷ Here, we make a higher level distinction between two very broad categories with distinctive types of creative activity which we will call Creative Services and Creative Content.⁴⁸ A full list of occupation and industry codes used in this report are given in Appendices 1 and 2 respectively.

Table 5.2.4 provides the totals for employment in these two broad categories, and Table 5.2.5 shows what these are as a proportion of total creative economy employment, for ease of comparison.

The creative service industries accounted for 1.2 times as many jobs as the creative content industries in the UK over the 2011-2013 period, while creative services occupations accounted for as many as 1.8 times the creative content occupations.

Interestingly, the small proportions of 'cross-employment' (creative service occupations in creative content industries, and vice versa),⁴⁹ suggest that, just as the creative industries are themselves a specialised employer of creative talent, each broad category specialises in a particular type of talent within the creative industries. Tables 5.2.4 and 5.2.5 show that in the main the creative content industries employ creative content workers and the creative service industries employ creative services workers.

5.3 Growth of the Creative Economy

Unfortunately, the classification changes in the SIC codes with the shift from SIC03 to SIC07, (implemented in the 2008 data), and in the SOC codes from SOC2000 to SOC2010 (implemented in the 2010 data) precluded a longer historical analysis at the time of writing this report (though an approximate series on a SOC2000 basis for the 2004-2010 period is provided in Dynamic Mapping).⁵⁰ Shortly before the publication the DCMS published a statistical release (DCMS 2015), containing the first backcasting of creative economy employment back to 1997.

Note: The average annual growth rate is the arithmetic mean of the growth rate in the two periods.

Table 5.3.1 shows the average annual growth rate over the period 2011-2013. It shows the creative economy grew 3.6 times faster than the UK workforce as a whole over this period.

Workforce share and its growth

Table 5.3.2 presents the evolution of the creative economy in terms of its contribution to the overall workforce.

5.4 Regional Geography of the UK's Creative Economy

Table 5.4.1 presents the size of the different components of the creative economy and the workforce as a whole for the UK's regions and Devolved National Administrations (DNAs),⁵¹ reported as the average of 2011 to 2013 to minimise fluctuations related to sample survey biases. These appear in descending order of the size of their creative economies' employment.

Note: There are a small number of people in the survey sample that count towards the UK totals in this and subsequent tables, but are not allocated to any region.

Figure 5.4.1 shows the regional/DNA distribution of the creative economy workforce, London and the South East account for a very significant share of creative economy employment, as noted previously in Freeman (2010). Together London and the South East of England region account for 43 per cent of employment in the creative economy in the UK.

Table 5.4.2 presents the different components of the creative economy as shares of the regional workforce. The particular importance of the creative economy to London's economy is apparent in this table. The creative economy's share of London's workforce - at 15.5 per cent - is almost double the national figure. Creative occupations in creative industries are also a higher proportion of the creative economy than in other regions, at 43 per cent (=6.6/15.5), a point consistent with earlier findings.⁵²

Workforce by Region and DNAs, UK, Average 2011-2013

Creative workforce shares may be helpfully analysed using location quotients (LQs). These are defined for the creative economy as the creative workforce share of the region (C_ER/W_FR) divided by the creative workforce share of the national workforce (C_EUK/W_FUK). As such, they allow us to compare how the importance of employment in particular occupations or industries in a region compares with their importance in the country as a whole. An LQ>1 means the regional workforce is more concentrated than the national one, an LQ=1 means that the concentration is the same and an LQ<1 means that it is less concentrated.

The table of location quotients is calculated by simply dividing every regional row of Table 5.4.2, by the corresponding UK row yielding Table 5.4.3.

A great deal has been written about the agglomeration economies enjoyed by London's creative industries (e.g. Freeman 2009, Pratt 2011, Chapain et al., 2010). Cities provide the combination of resources that allow the many different types of creative worker to interact with each other, on a face-to-face basis. This helps to explain why many creative industries form very tight clusters.⁵³ However, agglomeration alone is unlikely to explain why London is so distinctive in comparison to other large connected metropolitan agglomerations such as the West Midlands or the North West of England, each of which have more than half the population of London.⁵⁴

Rather, London's great attractiveness as a location for creative businesses is also likely to be due to its role as a 'world city'. The connectivity of world cities makes them, for example, the preferred location for gateway institutions that manage pre-market selection (that is, where creative content is selected by, for example, galleries, film producers, and publishers before going to market). In this way, world cities are the site of 'world galleries', festivals, and so on, which are not to be found in equivalent density in other large metropolitan areas which, on grounds of size alone, one might expect to be a focus of attraction for creative producers and their employees.⁵⁵

Agglomeration alone is also insufficient to explain the unevenness in the distribution of the creative economy and its components outside the Greater South East of England (the area consisting of London, the South East and the East of England regions), since all the regions listed contain quite major and connected urban concentrations. What explains, for example, the fact that the North East has a creative economy location quotient of only 0.59? We are examining the extent of agglomeration economies and spillovers more widely in new research.

Where is the creative economy growing?

Table 5.4.4 shows the annual average growth rates for employment in different components of the creative economy from 2011 to 2013, by region and DNA in the UK.

These tables show that the fastest growing parts of the UK's creative economy over this period have been the East of England, the West Midlands and the North East of England. With the exception of Scotland, which has experienced a negative growth rate on average in its creative economy over this period, London has been the slowest growing region.

Having analysed the size, growth and regional geography of the creative economy we present equivalent results for the high-tech economy.

6. The High-Tech Economy

In this section, we define the high-tech economy in terms of a set of high-tech industries and STEM occupations embedded in the wider economy, and from this derive estimates of its size, growth since 2011 and regional geography.

6.1 Defining the High-Tech Economy

In our approach we propose a baseline classification for STEM occupations based on a review of pre-existing treatments of STEM work. We compare the industries selected as STEM-intensive – or 'high-tech' – at different intensity thresholds with a classification of high-tech industries produced by the EU's statistical agency, Eurostat, based on R&D and knowledge intensity, and use this analysis to derive a classification that we use in this report.

6.1.1 Definition of STEM occupations in the literature

As discussed in Section 3, the concept of STEM has its origins in concerns that the qualifications of the workforce, and in particular, the supply of STEM-qualified graduates, do not match the needs of industry. There have been perennial complaints that the UK economy is not producing enough high-quality STEM graduates. There has, however, been remarkably little attempt to build a consensus on what constitutes a STEM occupation.

For this report we reviewed the detailed STEM occupation classifications employed in the studies listed in Table 6.1.1.1.

Note: In cases where the classifications are given in terms of the North American codes of the NAICS system, we have converted them to the nearest SOC2010 equivalents.⁶¹

According to the broadest classification we consider, that of the BLS (2010), it turns out that there are over four million STEM jobs in the UK. The narrowest, derived from Hecker, yields less than half this number, as shown in Table 6.1.1.2. The full list of occupations treated as STEM according to different accounts is given in Appendix 1, Table A1-4.

This table confirms that the considerable variation in how STEM occupations are classified has a significant effect on the number of jobs that are estimated to be in STEM occupations.

In the next section we undertake an analysis of STEM intensity according to these different STEM classifications, to determine whether a distinct set of high-tech industries can be identified that have an especially high percentage of STEM workers in their workforces.

6.1.2 How the distribution of STEM intensity is affected by different definitions of STEM occupations

To examine STEM intensity as a means of selecting a group of high-tech industries, we first construct the frequency distributions shown in Figure 6.1.2.1 for the set of STEM classifications given in Table 6.1.1.2. Each chart shows the distribution of STEM intensity for different classifications.

It is apparent from the chart that regardless of which of the four classifications of STEM occupations is used there is no one distinct set of industries where these occupations are uniquely unusually concentrated. This means that, unlike the case of creative intensity, we cannot use STEM intensity on its own with these definitions to identify the high-tech industries - that is, the employment of STEM occupations in the UK economy is much less specialised than is the case with creative workers.

How can we proceed? The distributions of STEM intensity above suggest that the discriminatory power of STEM intensity is perhaps greater for the narrower classifications of Hecker and RAE.

A theoretical justification for a narrower classification would be that STEM occupations constitute, in some way, one of the means by which scientific 'knowledge' finds its way into the processes and outputs of an enterprise or activity. Formally, we might think of them as:

roles explicitly involved in the development of new knowledge from the domains of science and engineering or in the rigorous application of existing science and engineering knowledge and processes to new circumstances.

The application of existing science and engineering knowledge and processes is necessary, but not sufficient according to this view, for an occupation to be a STEM occupation - though we accept this is a question of judgment. Routine application of engineering processes to standardised product or process should not be sufficient to be considered a STEM occupation. For this reason, occupations with high administrative or manual components - such as laboratory technicians or air-conditioning technicians are outside the scope of STEM occupations in our treatment.

Similar to creative occupations, STEM occupations involve weighing up alternatives and making decisions and choices that can significantly alter the outcome of the work. The routine following of a STEM-developed process, such as on a production line or in the installation of network cabling in an office does not in this conception make the role "STEM". For this reason, we do not classify as STEM many technician occupations, even when they are in a STEM field. Implementing someone else's STEM decision or checklist does not in itself, we argue, make the occupation itself STEM.

How does this view of STEM occupations compare with our treatment of creative work? We think STEM and creative occupations differ in at least two ways:

• Uniqueness Creative occupations are focussed on adding value through uniqueness - the market rewards those creators who can tap into the wants and needs of users, including businesses, with a product or service that is unique, even if only for a short time. Success in the fields of science and engineering knowledge is also initially measured through the novelty of the contribution, but subsequent to this the replicability of a result must always be established.
• Domain One of the criteria for creative occupations is that the output of the occupation is typically considered to be a creative service or product whatever the industry setting. In the case of STEM, we suggest, the range of output domains is much broader - from construction (material sciences), manufacturing (robotics), to agriculture (pesticides research and chemical engineering) and financial services (big data modelling).

By reviewing the set of STEM classifications used to generate the distributions in Figure 6.1.1 and applying the principles above, we propose a pragmatic set of STEM occupations, shown in Table 6.1.2.1 (described as 'baseline' to acknowledge the reality that classifying individual occupations as STEM or not STEM or indeed creative or non-creative - is inherently fuzzy).

TABLE 6.1.2.1 BASELINE STEM OCCUPATIONS

This selection is very close in spirit to that of Hecker (2005), with the major difference that we exclude a number of technician roles that Hecker includes.

6.1.3 Determining the STEM intensity threshold and identifying the high-tech industries

Figure 6.1.3.1 shows the frequency distribution of industry STEM intensities using our baseline STEM occupations. Although not bimodal, it suggests that - when combined with other indicators - STEM intensity is nonetheless a useful statistic for classifying a group of industries as high-tech.

To identify these other indicators we begin with an alternative approach to defining the high-tech industries which incorporates criteria that are independent of occupational intensity, or at least are not directly dependent on it.

This is the approach taken by Eurostat, which separately considers high-tech manufacturing and high-tech knowledge-intensive services. The group of industries selected is based on the NACE revision 1 classification, and is given in Table 6.1.3.1 (NACE is the EU's statistical classification of economic activities). ⁶², ⁶³ The Eurostat high-tech manufacturing classification is a NACE translation of the OECD's classification of high-tech manufacturing.⁶⁴

TABLE 6.1.3.1 HIGH-TECH INDUSTRIES AS DEFINED BY EUROSTAT IN NACE 1.1

Source: Eurostat, Aggregations of manufacturing and services based on NACE Rev 1.1

NACE industry classifications are generally equivalent to the previous version of the UK Standard Industry codes (SIC92 and SICO3), so we translate the NACE codes to these and then map them on to SICO7 classifications.⁶⁵

The following two tables show the industry classifications that are the most appropriate fit to the 2-digit NACE specified above - along with average employment over 2011 to 2013 and the STEM intensity calculated the same period.

TABLE 6.1.3.2 STEM INTENSITY OF HIGH-TECH KNOWLEDGE-INTENSIVE INDUSTRIES (BASED ON EUROSTAT DEFINITION), UK, AVERAGE 2011-2013

TABLE 6.1.3.3 STEM INTENSITY OF R&D-INTENSIVE INDUSTRIES (BASED ON EUROSTAT DEFINITION), UK, AVERAGE 2011-2013

As shown in Tables 6.1.3.2 and 6.1.3.3, a significant proportion - though by no means all - of the industries classified as high-tech in the Eurostat classification have high proportions of STEM employment relative to industries as a whole. In other words, STEM intensity appears to work relatively well in selecting those industries that are, on the grounds of their science and technology R&D spend or investment in other forms of knowledge, considered to be high-tech according to the Eurostat classification.

However, STEM intensity also selects a number of industries that are less R&D or knowledge-intensive. One way to view the different options is to consider the effect of varying the intensity threshold on the selection of STEM-intensive (high-tech) industries. This is shown in Figure 6.1.3.2. We use the term 'unanticipated negatives' to denote those industries that are identified as high-tech under the Eurostat classification but have a STEM intensity below the threshold STEM intensity, so would not be classified as high-tech using that threshold. 'Unanticipated positives' denotes those industries that are identified as high-tech as their STEM intensity is greater than the threshold, but are not classified as high-tech according to the Eurostat classification.

We adopt the following pragmatic procedure for selecting the threshold where we trade off the objective of minimising employment in the group of 'unanticipated positives' with the aim of capturing as many of the Eurostat high-tech industries as possible.

From Figure 6.1.3.2, it can be seen that as the STEM intensity threshold increases (i.e. the high-tech industry definition becomes more restrictive) the absolute number of people employed in industries that would be counted as high-tech under this definition (but which are not high-tech in the Eurostat definition) starts to fall off in discontinuous jumps until around the 15 per cent level where the effect of a higher rate of STEM intensity becomes more gradual, i.e. from this point onwards there are a set of smaller industries with STEM intensities clustered in the 15 per cent plus range.

If a lower STEM threshold of, say, 10 per cent was therefore adopted, we would be including significant numbers of jobs (over two million) in industries that are not considered as high-tech by Eurostat.

A 15 per cent threshold - as can be seen from Table 6.1.3.2 and 6.1.3.3 - would also mean that all but five of the industries in the Eurostat classification would be included (in contrast, a further six industries would drop out if a 20 per cent threshold intensity was instead used).⁶⁶ On this basis, we adopt a threshold of 15 per cent as the intensity threshold for our baseline high-tech industry classification.

As with the treatment of creative industries (and described in detail in Dynamic Mapping), following the application of the threshold intensity we remove those industry codes for which the APS sample size is too small to be statistically reliable (a threshold of 4,000 was used for this). This results in a baseline high-tech industrial classification, shown in the Table 6.1.3.4 below.

TABLE 6.1.3.4 HIGH-TECH INDUSTRIES (EMPLOYMENT AND STEM INTENSITY)

Having defined the UK's high-tech economy in terms of a given set of occupations (Table 6.1.2.1) and industries (Table 6.1.3.4), we can quantify its size, and growth and describe its regional geography.

6.2 THE SIZE OF THE HIGH-TECH ECONOMY

The preceding analysis suggests that whilst there are essential features of the high-tech economy that differ from the creative economy - in particular, that STEM workers have a less pronounced tendency to be employed in a specialised set of industries - there is nonetheless a strong analogy, because STEM occupations play a distinctive role in the production process.

The overall numbers of high-tech jobs in the UK are given in Table 6.2.1. The table shows that the UK's high-tech economy accounted for 3.2 million jobs in 2013, of which 2.4 million were in the high-tech industries. Around 50 per cent of STEM jobs were outside the high-tech industries in the wider high-tech economy, a strikingly similar degree of embeddedness to creative jobs.

TABLE 6.2.1 THE HIGH-TECH ECONOMY OF THE UK, 2011-2013

This, like the creative economy, can be displayed in the same trident format as Table 5.2.2, and is given in Table 6.2.2.

TABLE 6.2.2 HIGH-TECH TRIDENT AVERAGE 2011-2013

6.3 GROWTH OF THE HIGH-TECH ECONOMY

Table 6.3.1 shows the growth in employment within the various components of the UK's high-tech economy over time.

TABLE 6.3.1 EMPLOYMENT GROWTH IN THE MAIN HIGH-TECH ECONOMY COMPONENTS (PERCENTAGE CHANGE ON PREVIOUS YEAR)

The table shows that the high-tech economy grew by 2.1 per cent p.a. on average between 2011 and 2013, higher than the average rate of growth of the UK workforce (1.2 per cent p.a.). This growth was driven by STEM jobs (2.5 per cent p.a. in high-tech industries and 3.7 per cent p.a. outside), with growth in non-STEM roles in the high-tech industries being slower at 1 per cent p.a.

6.4 REGIONAL GEOGRAPHY OF THE UK'S HIGH-TECH ECONOMY

Table 6.4.1 presents the size of the different components of the high-tech economy and the workforce as a whole for the UK's regions and devolved national administrations, reported as the average of 2011 to 2013 to iron out year-on-year fluctuations. These are placed in descending order of the size of their high-tech economies' workforce. Figure 6.4.1 presents these numbers as contributions to the UK's high-tech economy workforce.

TABLE 6.4.1 HIGH-TECH ECONOMY EMPLOYMENT IN THE REGIONS AND DNAS, UK, AVERAGE 2011-2013

Note: There are a small number of people in the survey sample that count towards the UK totals in this and subsequent tables, but are not allocated to any region.

Figure 6.4.1 shows that London and the South East account for a very significant share of UK high-tech employment, though at 31 per cent this is less pronounced than in the case of the creative economy (43 per cent). Table 6.4.2 gives the regional workforce shares for all the components of the UK's high-tech economy. With a workforce share of 10.6 per cent, London has a smaller proportion of high-tech employment than the South East, and has a proportion equivalent to that of Scotland - strikingly different from the creative economy, where the difference between London and Scotland's creative economy workforce share is nine percentage points (Table 5.4.2).

Table 6.4.3 gives the corresponding location quotients (which as before normalise the various workforce shares for the region/DNA's overall share in the UK workforce) and for convenience gives the creative economy equivalents. This confirms the general finding that the distribution of high-tech economy jobs is more even than in the case of the creative economy, with location quotients ranging from 0.72 to 1.24 compared with those for regional creative economy which range from 0.59 to 1.86.

TABLE 6.4.2 GEOGRAPHICAL DISTRIBUTION OF THE COMPONENTS OF THE HIGH-TECH WORKFORCE SHARE BY UK REGIONS AND DNAS, AVERAGE 2011-2013

6.4 Location Quotients for the High-Tech Economy

Table 6.4.4 shows that the UK regions/DNAs with relatively high workforce shares in the high-tech economy also tend to have higher workforce shares in the creative economy (the rank correlation for the LQs is 0.85). This echoes a finding in Chapain et al., (2010) that there are strong patterns of geographical co-location between high-tech and creative firms (such as the co-location of advertising, software, computer games, and electronic publishing with high-tech manufacturing) in the UK. The regions which have LQs above 1 in both the high-tech and creative economies are the three neighbouring regions that constitute the greater South East of England.

KEY FINDINGS

• The South East of England and London account for 31 per cent of employment in the high-tech economy in the UK. However, as a share of the workforce in London, the high-tech economy accounts for fewer jobs than does the creative economy (10.6 per cent of the workforce compared to 15.5 per cent).
• The high-tech economy is in fact an important employer in all the regions and DNAs in the UK, not just the South of England. In other words, the high-tech economy is more evenly distributed compared with the creative economy, even though London, the South East and the East of England are significant regions in both cases.

Where is the high-tech economy growing?

Table 6.4.5 shows the average annual growth rate between 2011 and 2013 in employment across the regions and DNAs for each component of the UK's high-tech economy.

KEY FINDINGS

• The regions with the fastest growing high-tech economy over the 2011-2013 period have (in order) been Scotland, Yorkshire and Humberside, London and the West Midlands.
• The number of STEM jobs embedded in the wider economy fell in Northern Ireland, the South West of England and the East and West Midlands, over the period.

7 THE CREATIVE AND HIGH-TECH ECONOMIES COMPARED

How in an accounting sense do the creative and high-tech economies relate? To what extent do they intersect given the presence of occupational and industrial codes common to both of them? Are there any particular geographical patterns to activities which are both creative and high-tech?

Table A1.3 (Appendix 1), summarises the relationship between creative and STEM occupations in our analysis. Of the 22 STEM and 30 creative occupations, there are four common SOC codes (Information technology and telecommunications directors (1136); IT business analysts, architects and systems designers (2135); Programmers and software development professional (2136), and Web design and development professionals (2137)).

Table A2.3 (Appendix 2) summarises the relationship between the creative and high-tech industries. Three of the 31 creative industry codes are also classified as high-tech in our definition (Other software publishing (5829), Computer programming activities (6201) and Computer consultancy activities (6202)).

The coinciding occupations and industries are all Information and Communication Technologies (ICT) related occupations and industries. Arguably, this reflects the fact that ICTs are an example of what economic historians call a general purpose technology, playing a role analogous to electricity in the late 19th century - transforming the nature of all production in the economy.74 Instead of just separating out ICT industries, therefore, it makes sense also to acknowledge that both high-tech and creative industries make extended and universal use of ICT (even if that may be in different and distinctive ways).75

Figure 7.1 plots the relationship in accounting terms between STEM and creative occupations, as we have classified them, on a Venn diagram. It shows that there were around 0.46 million jobs that were both STEM and creative (average 2011-2013). Although these are ICT jobs it should be noted that this number is significantly smaller than broader treatments of ICT work e.g., that used by e-Skills UK in its 2013 statistical release on the UK's Information Economy which estimates that there are over one million people employed in ICT occupations in the UK.76

Tables 7.1 and 7.2 give the UK breakdown by region/DNA of the number and workforce share of jobs in creative and STEM occupations.

Note: There are a small number of people in the survey sample that count towards the UK totals in this and subsequent tables, but are not allocated to any region.

Jobs that are both creative and STEM are particularly prevalent in London and the South East, accounting for 2.3 per cent and 2.1 per cent of their overall workforces, respectively. Along with the East of England, these are the only regions where the proportion of employment is higher than the UK level of 1.5 per cent.

As shown in Figure 7.2 the number of jobs in industries that are both creative and high-tech is around 0.54 million. These are jobs in the three ICT-related industries mentioned earlier.

Tables 7.3 and 7.4 below give the UK breakdown by region/DNA of the number and workforce share of jobs in creative and high-tech industries.

Note: There are a small number of people that count towards the UK totals in this and subsequent tables, but are not allocated to any region.

Echoing the result for STEM occupations, this table shows that those industries that are classified as both creative and high-tech (the three ICT-related industries) are again particularly important employers in London (2.7 per cent) and the South East (2.8 per cent) and the East of England (2.1 per cent) compared with the rest of the country.

Figure 7.3 brings together Figures 7.1 and 7.2 at the UK level to show the intersection between the creative and high-tech economies (average 2011-2013). It shows that there were 0.87 million jobs in this segment.

Table 7.5 further shows how these jobs split into specialist employment, non-specialist workers and embedded roles.

Note: Numbers may not sum due to rounding

The table confirms that of the 0.87 million jobs that are located in both the creative and the high-tech economies, 0.46 million were in occupations that are both creative and STEM (the shaded areas in the table above). The cells outlined in bold are the 0.54 million jobs in industries that are both creative and high-tech.

Table 7.5 also shows that there were 20,000 STEM jobs (that were not in creative occupations) that were non-specialist jobs in creative industries that are not high-tech, and 58,000 creative jobs (that were not in STEM occupations) that were non-specialist jobs in high-tech industries that are not creative.

Growth in high-tech and creative employment

Table 7.6 re-emphasises the finding from earlier sections that the UK's creative economy has been growing over three times faster than the workforce as a whole at 4.3 per cent p.a. on average between 2011 and 2013, compared with a 2.1 per cent p.a. growth rate in the high-tech economy.

Table 7.7 reveals that the part of the UK workforce which straddles both the creative and high-tech economies (industries) has been particularly dynamic over this period, growing at 8.0 per cent (9.6 per cent) p.a.

KEY FINDINGS

• A core set of four ICT occupations account for the overlap between the 22 STEM and 30 creative occupations in our classifications. These are Information technology and telecommunications directors (1136); IT business analysts, architects and systems designers (2135); Programmers and software development professional (2136), and Web design and development professionals (2137).
• The number of jobs that are both creative and STEM is around 0.46 million. They account for 2.3 per cent, 2.1 per cent and 1.7 per cent of the London, South East and East of England workforces respectively, the only regions where the workforce share is higher than the national level of 1.5 per cent.
• Three of the 31 creative industry codes are also classified as high-tech in our definition: Other software publishing (5829), Computer programming activities (6201) and Computer consultancy activities (6202). Between them they account for 0.54 million jobs, although not all are in creative occupations.
• They account for 2.8 per cent, 2.7 per cent and 2.1 per cent of the South East, London and East of England workforces respectively, the only regions where the workforce share is higher than the national level of 1.8 per cent.
• 0.87 million people work in the intersection of the high-tech and creative economies i.e. either work in an occupation or industry that is in both the high-tech and creative economy.
• The part of the workforce which straddles both the creative and high-tech economies (industries) has been particularly dynamic over this period, growing at 8.0 per cent (9.6 per cent) p.a.

The next section examines the geography of the creative and high-tech economies at a sub-regional level.

8 SUB-REGIONAL GEOGRAPHY OF THE UK'S CREATIVE AND HIGH-TECH ECONOMIES

To understand the geography of the creative and high-tech economies, and to formulate supportive policies, it is necessary to analyse it at a sub-regional level. Analysis at a higher degree of spatial resolution allows concentrations of particular economic activity to be more accurately identified. Regions and DNAs can contain greater variations in economic characteristics within them (e.g. urban vs rural) than between them, so comparisons between regions/DNAs can potentially be misleading. Concentrations of creative and high-tech activity may also occur at small spatial scales, the visual effects industry cluster in London's Soho, for example, so it is desirable to analyse the data at the smallest possible scale for which official data are available.77

There are, however, limitations to analysing the Annual Population Survey (APS), or indeed any survey, for small units of geography: the sample sizes become smaller so that inferences on the characteristics of individual localities become less robust. The statistical authorities also impose restrictions on reporting numbers due to them being potentially disclosive of survey participants' identities.

In this report, we adopt the sub-regional geography of the European Union's NUTS (Nomenclature of Territorial Units of Statistics) classification.78 This is the geographic system used by the European Union and has the benefit of allowing international comparisons.79 There are three spatial levels in the NUTS system: NUTS1 is the largest geography and corresponds to UK regions, NUTS2 is the second largest geography corresponding to counties or groups of counties, and NUTS3 is the geography with the smallest spatial scale. NUTS3 units can be aggregated to create NUTS2 and NUTS1 geographies and correspond to a mixture of UK counties and unitary authorities/urban areas. As the APS is based on place of residence, this means, however, that in some cases respondents will be working in an area different from where they live i.e. employment allocated to a given NUTS3 geography may be located elsewhere.

In the following set of maps we plot the location quotients i.e. the proportion of the area's workforce in a given activity relative to the share of that activity in the UK workforce as a whole.80 This is undertaken for the creative economy and its components, followed by the high-tech economy. Areas that are more darkly shaded in the maps are those where there is a higher proportion of employment in creative and high-tech activity relative to the national level (i.e. those with a higher location quotient). As Northern Ireland is not coded at the NUTS3 level within the APS it is plotted at the NUTS1 level i.e. as Northern Ireland. This means concentrations of high-tech or creative activity at lower levels of spatial resolution in Northern Ireland will not be identified.

8.1 The sub-regional geography of the UK's creative economy and its components

Figure 8.1.1 shows that creative specialists (i.e. people in creative occupations working in creative industries) are a particularly high proportion of employment relative to their proportion in the UK workforce in West Inner London, East Inner London, West and North West Outer London, South Outer London and Brighton & Hove. They are also particularly prevalent in the counties to the west (Berkshire, Oxfordshire, and Buckinghamshire), north (Hertfordshire and Cambridgeshire) and south (Surrey) of London, and in Bristol, Edinburgh and Glasgow.

The creative industries account for a particularly high workforce share in East Inner London, West Inner London and Brighton & Hove. They are also a significant proportion of the workforce in West and North West Outer London, South Outer London,84 to London's west (Berkshire, Oxfordshire, Buckinghamshire) the immediate north (Hertfordshire and Cambridgeshire) and south (Surrey), and in Bristol and Edinburgh.

The wider creative economy is a particularly significant employer as a proportion of the workforce in East Inner London, West Inner London, Surrey, Brighton & Hove and Outer London (West and North West, and South). It is also concentrated to the west (Berkshire, Oxfordshire, Buckinghamshire, Milton Keynes), the north (Hertfordshire and Cambridgeshire) and in Edinburgh and Bristol.

8.2 The sub-regional geography of the UK's high-tech economy and its components

Figure 8.2.1 shows that STEM occupations in high-tech industries account for a particularly high share of employment relative to their share of the UK workforce in Berkshire (largest towns: Reading and Slough) and in Cambridgeshire. They also account for a high proportion of the workforce in the counties to the north, west and south west of London, Outer London (West and North West) and Cheshire, Gloucestershire, North and North East Somerset, Aberdeen City & Aberdeenshire, Halton and Warrington and Bedfordshire. Urban areas with a particularly high proportion of STEM occupations working in high-tech industries in the workforce include: Glasgow, Bristol, Milton Keynes, Swindon, and Edinburgh.

Employment in high-tech industries as a proportion of areas' workforces is more evenly distributed, being also relatively prominent in West Cumbria, Flintshire and Wrexham, Derby, South and West Derbyshire, Caithness and Sutherland, Portsmouth, Hartlepool and Stockton-on-Tees, and Warwickshire.

Some of the agglomerations are likely to relate to the supply chains and employment of specific large-scale employers: for instance, Rolls Royce plc employs its largest concentration of aeronautic engineers in the UK in Derby.85 In West Cumbria, there are two major employers in the form of the Sellafield Nuclear site (the largest nuclear site in Europe86) and BAE systems' submarine operations in Barrow-in-Furness.87 Aberdeen City and Aberdeenshire – the area with the highest LQ – is the centre of the UK's North Sea oil industry. The pattern of concentrations of high-tech activity also relate to the transport links along the M4 corridor and the close links with Heathrow airport which have stimulated technology in this area with a number of prominent software companies locating nearby, for example Microsoft and Oracle which have their UK headquarters in Reading.

The high-tech economy represents a particularly high proportion of the workforce in Aberdeen City & Aberdeenshire, the counties to the north, west and south west of London, and in Cheshire, Warwickshire, Halton and Warrington and West Cumbria. Urban areas with a particularly high proportion of high-tech economy employment include: Milton Keynes, Derby, Swindon, Edinburgh, Outer London (West and North West) and Bristol.

The lower intensity threshold (15 per cent as opposed to 30 per cent in the case of the creative industries) for being in a high-tech industry has the consequence that non-specialist/support workers account for a larger proportion of the workforce than in the case of the creative industries, so the change in geographical distributions when comparing STEM employment in the high-tech industries with high-tech industry employment may, depending on the industries involved, be larger than in the case of the creative industries. This helps to explain the more marked differences between the high-tech economy and high-tech industries maps compared with the differences between the creative economy and creative industries maps.

8.3 How the sub-regional distributions of the creative and high-tech economies compare

Tables 8.3.1 and 8.3.2 show the location quotients for the high-tech and creative economies and their components sorted by the 20 NUTS3 areas with the largest creative and high-tech economy location quotients respectively.

It is immediately apparent from these tables that the creative economy is especially highly concentrated in Inner London in a way that the high-tech economy is not (in fact, Inner London does not even feature in the top 20 location quotients for high-tech).

We might reasonably speculate that this pattern is explained by the fact that high-tech manufacturing industries are in the main more capital-intensive than creative industries, and therefore require more space to operate than is economical in central London. This is less likely to be the case for many digital media and software businesses which are at the nexus of the high-tech and creative economies, such as those that have amassed in Tech City.88 Indeed, it has been argued that such firms are increasingly locating in city centres due to their attractiveness of those locations to the talent they employ, the need to draw on the range of different skills that cities offer, and technological developments, such as cheap cloud storage, which is allowing companies to operate with smaller footprints.89 If we examine the LQs of the intersection of the high-tech and creative economy then the LQ for Inner London East becomes much higher at 1.5, which is consistent with this (although this is not the highest LQ in London). A recent study (KPMG Data Monitor, 2013) using Annual Business Register and Employment Survey (BRES) data, at a lower spatial of resolution (the London borough level) also did not find the tech-cluster in the Old Street area to be that pronounced relative to others in London.90 That study explained this as being due to the London boroughs of Hackney and Islington, having substantial and diverse economies containing many multinational companies, particularly in the financial services, advertising and market research industries. In the case of that study it also considered that the BRES data was likely to under-record start-up activity as it is based on VAT registrations that will omit smaller companies. In the case of the current study the use of APS data, which is based on where workers reside, may also be making it harder to distinguish the existence of the East London tech cluster.

Another location that one might expect to appear in the above list, but which is absent is Manchester (the relevant NUTS3 geographies in this case being Greater Manchester – North, and Greater Manchester – South). These employ a large number of people in the creative economy, with both areas in the top quartile of NUTS3 areas in employment terms. Their absence is partly explained by the fact that London is so prominent in the creative economy thus having a substantive effect on the proportion of those employed in the creative economy in the UK. If London is excluded from the calculation of location quotients, then the LQs of some of the Manchester areas become larger i.e. there is a high proportion of creative employment relative to the rest of the UK, but this is partly hidden by the amount of creative employment in London.

As many as 15 of the 20 areas with the highest location quotients for the creative economy are in London and the South East/East of England area, with the equivalent number for the high-tech economy being ten.

11 of the areas make both top-20 lists: Berkshire, Cambridgeshire, Oxfordshire, Buckinghamshire, Milton Keynes, Surrey, Hertfordshire, Hampshire, and Outer London (West and North-West), and outside the South East of England, Edinburgh and Bristol.

Table 8.3.3 provides further detail on the specialisation profile of different areas, showing areas with high-tech or creative economy LQs greater than 1.2 in one or both of the categories.

From this we can see a number of areas where there are particularly high concentrations of creative and high-tech employment as a proportion of an area's workforce relative to their respective shares of the UK workforce:

Of high-tech and creative activity: Berkshire, Cambridgeshire, Oxfordshire, Buckinghamshire, Surrey, Hertfordshire, Milton Keynes and Edinburgh, Bristol, West and North West Outer London.

Of high-tech only: Aberdeen City & Aberdeenshire, Derby, West Cumbria, Hampshire, Cheshire, Halton and Warrington, Swindon, Bedfordshire and Warwickshire.

Of creative only: East Inner London and West Inner London, South Outer London, and Brighton & Hove.

How geographically dispersed are the creative industries compared to the high-tech industries?

To analyse how geographically distributed high-tech and creative industries are in the UK compared with other sectors, Table 8.3.4 shows how they compare across three measures of the dispersion with the location quotients for the broad industrial groups used by the ONS.91

This confirms that employment in sectors like construction, distribution, transport, hotels and restaurants and government, health and education, where demand is broad-based on a geographical basis are evenly distributed across the country. In other sectors like agriculture, which are strongly driven by proximity to specific natural resources, industries are highly spatially concentrated. All three dispersion measures present a consistent story. Namely that:

• The high-tech industries are more evenly distributed across the UK than the creative industries.
• Financial and insurance industries (see also Figure 8.3.5) are even more highly concentrated than the creative industries, being especially important employers in central London and in areas like Edinburgh, Essex and Southend on Sea. The LQ for the financial and insurance industries in West Inner London is 3.18, the highest LQ of any sector anywhere outside of agriculture.

Source: APS 2011-2013 Note: The Arts and other sector. Although this category includes a number of Arts related activities (e.g. performing arts employment) it also includes employment relating to sporting activities, membership organisations and personal services e.g. washing and (dry-)cleaning of textile and fur products. The measure is therefore not comparable to the creative industries figure.

KEY FINDINGS

• The creative economy is an especially important employer in Inner London and in the counties surrounding London. While the creative industries are relatively geographically concentrated across the UK compared with many other sectors, they are less concentrated than sectors like financial and insurance services, and agriculture.
• The high-tech economy, although found in many of the same areas as the UK's creative economy, is more evenly distributed around the country. This partly reflects the fact that it contains a number of industries that are centred outside London and the South East.
• The colocation of the creative and high-tech economies and its concentration in the South East of England is not wholly explained by the ICT-related occupations and industries that commonly make them up. It also reflects a tendency for creative activities and high-tech activities more generally to collocate (echoing the earlier finding by Chapain et al, 2010).
• In the European NUTS3 sub-regional geography:
  1. 1Counties identified as being those where the creative and high-tech economies account for a particularly high proportion of jobs are:
     • High-tech and creative: Berkshire, Cambridgeshire, Oxfordshire, Buckinghamshire, Hertfordshire and Surrey.
     • High-tech only: Aberdeenshire & Aberdeen City, Cheshire, Hampshire, West Cumbria and Bedfordshire and Warwickshire.
  2. 2Unitary authorities/Urban areas identified as being where the creative and high-tech economies both account for a particularly high proportion of jobs are:
     • High-tech and creative: Bristol, Edinburgh, Outer London (West and North West) and Milton Keynes.
     • High-tech only: Derby, Swindon, Halton and Warrington.
     • Creative only: Inner London, South Outer London and Brighton & Hove.

9 CONCLUSIONS AND NEXT STEPS

CONCLUSIONS

Size of the creative and high-tech economies

• Creative There were 2.6 million jobs in the UK's creative economy in 2013. This consisted of 1.71 million jobs in the creative industries (890,000 in creative occupations and 818,000 working in other roles) and 907,000 jobs in creative occupations outside of the creative industries.
• High-tech There were 3.2 million jobs in the UK's high-tech economy in 2013. This consisted of around 2.4 million jobs in high-tech industries (825,000 in STEM (Science, Technology, Engineering and Mathematics) occupations and approximately 1.6 million in other roles) and 806,000 jobs in STEM occupations outside of the high-tech industries.

Growth of the creative and high-tech economies

• Creative Employment in the creative economy on average grew at 4.3 per cent p.a. 3.6 times faster than the workforce as a whole (1.2 per cent p.a.) between 2011 and 2013. Employment in creative industries grew faster still, at over four times the rate of the workforce. This was driven by strong growth in both creative occupations employed in creative industries (4.3 per cent p.a.) and in those employed in other occupations in creative industries (5.8 per cent p.a.).
• High-tech At 2.1 per cent p.a., employment in the high-tech economy also grew faster than the workforce as a whole between 2011 and 2013. High-tech industry employment grew on average by 1.5 per cent p.a., more slowly than the rate of growth in STEM occupations of 3.1 per cent p.a. (Within this the rate of growth of STEM occupations outside the high-tech industries (3.7 per cent p.a.) was faster than that for those employed in high-tech industries (2.5 per cent p.a.)).

The intersection of the creative and high-tech economies

The figures below are the average of 2011-2013 data

• There were on average 0.46 million jobs in dual creative-STEM occupations in 2011-2013, accounting for 2.3 per cent of London's workforce, 2.1 per cent of the workforce in the South-East, and 1.7 per cent of the workforce in the East of England respectively. These were the only regions where the proportion of employment was higher than the level in the UK as a whole of 1.5 per cent.
• There were 0.54 million jobs in industries that were both creative and high-tech. They accounted for 2.8 per cent of the workforce in the South East of England, 2.7 per cent of the workforce in London and 2.1 per cent of the workforce in the East of England.
• 0.87 million jobs fell within both the high-tech and creative economies.
• This part of the workforce has been particularly fast-growing, at 8.0 per cent p.a. on average over the period 2011-2013. Within this, those ICT-related industries that can be considered as both creative and high-tech were particularly dynamic, growing at 9.6 per cent p.a.

Regional geography of the creative and high-tech economies

The high-tech and creative economies are disproportionately important employers in the Greater South East of England (London, East of England and South East) relative to the share that these sectors are of the UK workforce. The significance of London is even more pronounced in the case of the UK's creative economy than it is in the high-tech economy.

Interestingly, in recent years the creative economy employment has - Scotland's aside - been growing faster in regions outside London than it has in the capital itself.

Creative

• London and the South East of England account for 43 per cent of employment in the UK's creative economy.
• The creative economy employs 15.5 per cent of London's workforce, 10.7 per cent of the South East's workforce, and 8.4 per cent of the East of England workforce. Previous research suggests that the prominence of London is likely explained by strong agglomeration effects associated with London's size and with its status as a world city.
• The creative economy has been growing most rapidly in areas outside London between 2011-2013, particularly in the East of England, West Midlands, North East, and Yorkshire and the Humberside. Scotland is the only area where the creative economy has shrunk.

High-tech

• The South East of England and London account for 31 per cent of employment in the UK's high-tech economy.
• The high-tech economy employs 12.8 per cent of the South East's workforce and 11.2 per cent of the East of England's workforce. In contrast to the creative economy, this measure of the importance of high-tech to London's workforce is slightly behind Scotland, at 10.6 per cent compared with 10.9 per cent respectively.
• The regions/DNAs with the fastest growing employment in the high-tech economy over the 2011-2013 periods are: Scotland, Yorkshire and Humberside, London and the West Midlands.

Sub-regional geography of the high-tech and creative economy

In the European NUTS3 sub-regional geography:

• Counties identified as being those where the high-tech and creative economies are particularly important employers as a proportion of the geographies' workforce are:
  • High-tech and creative: Berkshire, Cambridgeshire, Oxfordshire, Buckinghamshire, Hertfordshire and Surrey.
  • High-tech only: Aberdeenshire & Aberdeen City, Cheshire, Hampshire, West Cumbria, Bedfordshire and Warwickshire
• Unitary authorities/Urban areas identified as being where the high-tech and creative economies are particularly important employers as a proportion of the geographies' workforce are:
  • High-tech and creative: Bristol, Edinburgh, Outer London (West and North West) and Milton Keynes.
  • High-tech only: Derby, Swindon and Halton & Warrington.
  • Creative only: Inner London (East and West), South Outer London, Brighton & Hove.

DISCUSSION AND NEXT STEPS

• The creative and high-tech industries have captured the imagination of policymakers in describing segments of the economy which are believed to display high levels of job creation, innovation and economic growth. However, conceptual ambiguities and the insufficient regard to transparent classifications that has beset these areas have hindered the design and evaluation of effective policies. Policymakers have been seemingly happy to use the terms 'industries' and 'economy' interchangeably, and employ inconsistent classifications that have made it impossible to track progress across key parts of the economy and over time.
• The definitional uncertainties are great, meaning that sensitivity analysis is critical,92 but in the baseline classifications adopted in this report employment in both the creative and high-tech economies since 2010 has been rapid, and much faster than in the UK workforce as a whole. Classification changes from SOC2000 to SOC2010 in the occupational codes precluded a historical analysis going back further in time during the writing of this report. The Office for National Statistics should give greater attention to how longer time-series for industrial and occupational analysis can be constructed when classification systems change ( ideally this would include back coding a sample of individual survey responses using revised codes).
• The APS data we have employed in this study has enabled us to reveal the size, growth and geography of the UK's creative and high-tech economies. In further research we are using the data to investigate the extent of possible knowledge spillovers from creative/high-tech activities on those working in other fields.93 Insofar as knowledge spillovers are associated with sub-optimally low levels of investment in knowledge - that is, they give rise to a market or systems failure - this may justify some form of policy intervention.
• Despite its large sample size, the APS is however a sample survey which limits its use at microspatial units of geography. The Household Census in principle provides a solution to this problem, but in practice is of limited use for our analysis. First, because the data is only available for 200 industry sectors at a mixture of 2-, 3- and 4-digit levels (this compares with the SICO7 standard of 617 codes available at 4 digits, of which 603 are used in the APS), which places major constraints on the ability of researchers to identify creative and high-tech industries in the Census data. Second, we cannot undertake regional analysis on Census data as the ONS has not released 4-digit occupational data within the 3-digit industry employment tables released at regional or sub-regional level. These restrictions severely limit the use of Census data for industrial policy analysis (a marked contrast with the US where the Household Census has been used in a multitude of influential studies of occupation and industry dynamics), and should be addressed by the ONS in its future work on the Census.

10 GLOSSARY

Creative Economy Those employed in creative industries (either in creative occupations or other roles) and those employed in creative occupations outside the creative industries.

Creative Industries Industries defined as creative under the Department for Culture Media and Sport (DCMS) definitions. These are set out in Appendix 2 (Table A2.1).

Creative Non-specialist Someone employed in the creative industries in an occupation which is not creative on the definition below.

Creative Occupations Occupations defined as being creative under the set of SOC codes in the Department for Culture Media and Sport definition set out in Appendix 1 (Table A1.1).

Creative Specialist Someone employed in a creative industry in a creative occupation.

Crosswalk A mapping between a set of occupational (or industrial classifications) done on a different basis e.g. SIC2007 and SIC2003.

DM, Dynamic Mapping The report which introduced the Dynamic Mapping approach. Bakhshi, H., Freeman, A. and Higgs, P. (2013) 'A Dynamic Mapping of the UK's Creative Industries.' London: Nesta.

Embedded Creative Someone employed in a creative occupation outside of the creative industries.

Embedded STEM worker Someone employed in a STEM occupation outside of the high-tech industries. STEM occupations being defined below.

High-tech Economy Those employed in high-tech industries (either in STEM occupations or other roles) and those employed in STEM occupations outside of the high-tech industries.

High-tech Industries Industries which have a STEM occupation intensity in excess of a threshold of 15 per cent, subject to passing the 4,000 employment robustness test (in our baseline classification). Set out in Appendix 2 (Table A2.2.).

High-tech Non-specialist Someone employed in a high-tech industry who is not in a STEM occupation.

Microspatial Clustering Small-scale clusters that form at spatial levels that may be below existing standardised geographies, for example a block of streets.

The Standard Industrial Classification (SIC) This is a means of classifying businesses according to the type of economic activity that they are engaged in. The latest version of SIC codes in the UK is SIC 2007, which superseded SIC 2003.

The Standard Occupational Classification (SOC) This is a means of classifying the occupation of a person according to the work they do and the skill level required.

STEM occupation An occupation that involves the use of Science, Technology, Engineering and Mathematics (STEM). The STEM occupational definition used in this report are set out in Appendix 1 (Table A1.2).

STEM Specialist Someone who is working in a STEM occupation in a high-tech industry.

Creative/STEM intensity The proportion of industry employment that is in the set of occupations considered as creative or STEM.

11. REFERENCES

Autor, D., Levy, F. and Murnane, R.J. (2003) The skill content of recent technological change: an empirical exploration. 'Quarterly Journal of Economics.' 118(4), November 2003.

Bakhshi, H., Freeman, A. and Higgs, P. (2013) ‘A Dynamic Mapping of the UK's Creative Industries' London: Nesta.

Bakhshi, H., Hargreaves, I. and Mateos-Garcia, J. (2013) 'A Manifesto for the Creative Economy.' London: Nesta.

Bakhshi, H., Lee, N. and Mateos-Garcia, J. (2013) 'Capital of Culture? An econometric analysis of the relationship between arts and cultural clusters, wages and the creative economy in English cities.' In Rushton (ed.) 'Creative Communities: Art Works in Economic Development.' Washington DC: Brookings Institution Press.

Bay Area Council Economic Institute Report (2012) 'Technology Works: High-Tech Employment and Wages in the United States.' San Francisco CA: Bay Area Council Economic Institute.

Beaverstock, J.V., Smith, R.G. and Taylor, P.J. (1999) A Roster of World Cities. 'Cities.' Vol. 16, Number 6, December 1999, pp. 445-458(14). Elsevier.

BIS (2012) 'BIS industrial strategy: UK Sector analysis.' London: BIS.

BIS (2011) 'STEM graduates in non-STEM jobs.' London: BIS.

BLS (2010) 'Options for defining STEM (Science, Technology, Engineering, and Mathematics) occupations under the 2010 Standard Occupational Classification (SOC) system: SOC Policy Committee recommendation to the Office of Management and Budget (OMB).' Washington DC: Bureau of Labor Statistics.

BOP (2014) 'World Cities Culture Report.' London: Burns Owen Partnership.

Brinkley, I. (2006) 'Defining the knowledge economy.' London: The Work Foundation.

Caves, R. E. (2002) 'Creative Industries: Contracts Between Art and Commerce.' Cambridge, MA: Harvard University Press.

Chapain et al., (2010) 'Creative clusters and innovation.' London: NESTA.

Chapple, K., Markusen, A., Schrock, G. and Yamamoto, Y. (2004) Gauging metropolitan "high-tech" and "I-tech" activity. 'Economic Development Quarterly.' 18 (1).

Cox, G. (2005) 'The Cox Review of Creativity in Business: Building on the UK's Strengths.' HM Treasury e-Comms Team.

Cunningham, S. (2014) 'Creative Labour, Creative Services, and Embedded Creatives.' In Hearn, G., Bridgstock, R., Goldsmith, B. and Rodgers, J. (2014) 'Creative Work beyond the Creative Industries: Innovation, Employment, and Education.' Cheltenham: Edward Elgar.

Currid, E. and Stolarick, K. (2010) 'The Occupation-Industry Mismatch: New Trajectories for Regional Cluster Analysis and Implications for Economic Development. ‘Urban Studies.' 47(2).

DCMS (1998) 'Creative Industries Mapping Document 1998.' London: DCMS.

DCMS (2001) 'Creative Industries Mapping Document 1998.' London: DCMS.

DCMS (2010) 'December 2011 Creative Industries Economic Estimates (Experimental).' London: DCMS.

DCMS (2011) 'December 2011 Creative Industries Economic Estimates.' London: DCMS.

DCMS (2014) 'Creative Industries Economic Estimates, January', London: DCMS.

DCMS (2015) 'Creative Industries Economic Estimates, January', London: DCMS.

Deroin, V. (2011) 'European Statistical Works on Culture: ESSnet-Culture Final report, 2009-2011.'

DFES (2006) 'The Supply And Demand for Science, Technology, Engineering And Mathematics Skills In The UK Economy.' London: DFES.

DIUS (2009) 'The Demand for Science, Technology, Engineering and Mathematics (STEM) Skills.' London: Department for Innovation Universities and Skills.

DTI (2006) 'Science, Engineering and Technology skills in the UK.' London: DTI.

Dyson, J. (2010) 'Ingenious Britain. Making the UK the leading High-tech exporter in Europe.'

Echeverri-Carroll, E. and Ayala, S. (2009) Wage differentials and the spatial concentration of high-technology industries. 'Papers in Regional Science.' Vol. 88, Issue 3, pages 623-641.

e-Skills UK, Intellect, BCS (2013) ‘Information Economy Economic estimates 2013'.

Eurostat, Aggregations of manufacturing and services based on NACE Rev 1.1.

Falk, R., Bakhshi, H., Falk, M., Geiger, W., Karr, S., Keppel, C., Leo, H. and Spitzlinger, R. (2011) 'Innovation and Competitiveness of the Creative Industries.' Vienna: Austrian Institute of Economic Research.

Florida, R. (2002) 'The Rise of the Creative Class.' New York: Basic Books.

Freeman, A. (2004) 'London's Creative Sector. 2004 Update.' London: Greater London Authority.

Freeman, A. (2008a) 'Culture, Creativity and Innovation in the Internet Age.' Presented to the conference on IPR, Birkbeck College, May 2008.

Freeman, A. (2008b) 'Benchmarking and Understanding London's Cultural and Creative Industries.' Presented to the conference of the Canadian Conference Board on Creative Industries, March 2008.

Freeman, A. (2009) 'Creative Industries 2009 Update.' London: Greater London Authority.

Freeman, A. and Cheshire, P. (2006) 'Defining and Measuring Metropolitan Regions: a rationale.' Paris: OECD.

Freeman, A. and Urwin, P. (2003) 'The GLA's London Workforce Employment Series.' London: Greater London Authority.

Frey, C. B. and Osborne, M.A. (2013) 'The Future of Employment: How Susceptible Are Jobs To Computerisation?' Oxford: Oxford Martin School, Programme on the Impacts of Future Technology, University of Oxford.

Garnham, N. (2005) From Cultural to Creative Industries: An analysis of the implications of the "creative industries" approach to arts and media policymaking in the United Kingdom. 'International Journal of Cultural Policy.' 11(1).

Gibbon, C. (2011) 'How much can the creative industries contribute to regional development in Britain?' London: Burns Owen Partnership.

Goodridge, P., Haskel, J. and Wallis, G. (2012), 'UK Innovation Index: Productivity and Growth in UK Industries.' Nesta Working Paper 12/09. London: Nesta.

Gordon, J. C. and Beilby-Orrin, H. (2001) ‘International Measurement of the Economic and Social Importance of Culture.' Paris: Statistics Directorate, OECD.

Government Office for Science (2010) 'Technology and Innovation Futures: UK Growth Opportunities for the 2020s'. London: Government Office for Science.

Growth Analysis (2009) 'Cultural industries in Swedish statistics: proposal on delimitation for future mappings.' Ostersund: Growth Analysis.

Greenwood, C., Harrison, M. and Vignoles, A. (2011) 'The labour market value of STEM qualifications and occupations.' London: Institute of Education.

Hadlock, P., Hecker, D. and Gannon, J. (1991) High technology employment: another view. 'Monthly Labor Review.' July 1991, pp. 26-30.

Hecker, D.E. (2005) High-technology employment: a NAICS-based update. 'Monthly Labour Review.' U.S. Bureau of Labor Statistics.

Hecker, D.E. (1999) High-technology employment: a broader view. 'Monthly Labor Review.' June 1999, pp. 18-28. U.S. Bureau of Labor Statistics.

Higgs, P., Cunningham, S. and Bakhshi, H. (2008) 'Beyond the Creative Industries: Mapping the Creative Economy in the United Kingdom.' London: NESTA.

Higgs, P., Cunningham, S., Hearn, G., Adkins, B. and Barnett, K. (2005) ‘The Ecology of Queensland Design.' Technical Report, CIRAC. Brisbane: Queensland University of Technology.

HM Treasury and BIS (2011) 'The plan for growth'. London: HM Treasury and BIS.

HM Government (2013) 'Information Economy Strategy'. London: HM Government.

ISIC (2002) 'International Standard Industrial Classification of All Economic Activities (ISIC) Revision 3.1.' New York: United Nations.

KPMG and Tech Monitor UK (2013) 'Understanding tech clusters and tracking the UK tech sector's outlook for employment and growth.'

Luker, W. and Lyons, D. (1997) Employment shifts in high-technology industries, 1988-96. 'Monthly Labor Review.' June 1997, pp. 12-25.

Mateos-Garcia, J., Bakhshi, H. and Lenel, M. (2014) 'A Map of the UK Games Indusrty.' London: Nesta and Ukie.

McLuhan, M. (1964) 'Understanding Media: The Extensions of Man.' New York NY: McGraw-Hill.

Moretti, E. (2013) ‘The New Geography of Jobs.' Boston MA: Mariner.

Nathan, M. and Vandore, E. (2014) Here Be Startups: Exploring a young digital cluster in Inner East London. Forthcoming in 'Environment and Planning.'

NSF (1988) 'Science and Technology Resources in U.S. Industry, special report.' Arlington, VA: National Science Foundation. pp. 88-231.

OECD (1996) 'The Knowledge-based Economy.' Paris: OECD.

OECD (2002) 'Frascati Manual: Proposed Standard Practice for Surveys on Research and Experimental Development.' Paris: OECD.

OECD (2005) 'Oslo Manual: The Measurement of Science and Technological Activities, Proposed Guidelines for Collecting and Interpreting Technological Innovation Data.' Paris: OECD.

OECD Directorate for Science, Technology and Industry (2011) ‘ISIC rev.3 Technology intensity definition.' Paris: OECD.

OHIM (2013) 'Intellectual property rights intensive industries: contribution to economic performance and employment in the European Union. Industry-Level Analysis Report.' Brussels: Office for Harmonization in the Internal Market.

Office of Technology Policy (2000) 'The Dynamics of Technology-Based Economic Development, State Science and Technology Indicators.' 4th ed. Washington DC: U.S. Department of Commerce.

ONS (2007) 'The Standard Industrial Classification 2007.' (UK SIC 2007). Newport: ONS.

ONS (2010) 'Standard Occupational Classification 2010.' (UK SOC 2010). Newport: ONS.

ONS (2011) 'Labour Force Survey User Guide volume 1 - LFS background and methodology'. Newport: ONS.

Perez, C. (2010) 'Technological Revolutions and Financial Capital: The Dynamics of Bubbles and Golden Ages.' Cheltenham: Edward Elgar.

Riche, R. W., Hecker, D.E. and Burgan, J.U. (1983) High technology today and tomorrow: a small slice of the employment pie. 'Monthly Labor Review.' November, 1983, pp. 50-58.

Rothwell, J. (2013) 'The hidden STEM economy.' Washington DC: Brooking Institution Metropolitan Program.

Santos Cruz, S. and Teixeira, A. (2012) 'Methodological approaches for measuring the creative employment: a critical appraisal with an application to Portugal.' FEP Working Papers No. 455.

Stoneman, P. (2010) 'Soft Innovation: Economics, Product Aesthetics, and the Creative Industries.' Oxford: Oxford University Press.

UKCES (2011) 'The Supply of and demand for High-Level STEM skills.' and UKCES (2013) 'Supply of and demand for High-Level STEM skills.' London: UKCES.

UKCES (2013) 'Supply of and demand for High-Level STEM skills.' London: UKCES.

UNESCO (2009) 'Framework for Cultural Statistics.' UNESCO Institute for Statistics. Paris: UNESCO.

United Nations (2010) 'Creative Economy Report 2010.' Geneva: United Nations.

U.S. Congress, Office of Technology Assessment (1982) 'Technology, Innovation, and Regional Economic Development.' Washington DC: U.S. Congress, Office of Technology Assessment.

USPTO (2012) 'Intellectual Property and the U.S. Economy: Industries in Focus.' Alexandria VA: USPTO.

Westlake, S. (2013) Rebalancing Act: Rationales and Policies for Sectoral Economic Rebalancing. 'Oxford Review of Economic Policy.'

Willets, D. (2012) 'Our High-tech future.' speech.

Willets, D. (2013) 'Eight great technologies.' Speech at Policy Exchange.

WIPO (2003) Guide on Surveying the Economic Contribution of the Copyright-Based Industries. Geneva: WIPO.

WIPO (2004) 'The Economic Contribution of Copyright-Based Industries in the USA.' Creative Industries Series No. 1. Geneva: WIPO.

APPENDICES

1. APPENDIX OF SOC TABLES

TABLE A1.1 SOC2010 CREATIVE OCCUPATIONS AND THEIR SEGMENTS

Creative Services Occupations

TABLE A1.2 SOC2010 STEM OCCUPATIONS

Engineering Occupations

IT Occupations

Science Occupations

TABLE A1.3 RELATIONSHIP BETWEEN CREATIVE AND STEM OCCUPATIONS

CREATIVE OCCUPATIONS

Creative Services Occupations

STEM Occupations outside the Creative Occupation definition (18 in total)

TABLE A1.4 SOC2010 CODES CORRESPONDING TO STEM OCCUPATIONS FOR VARIOUS CLASSIFICATIONS

Please note that the original Hecker, BLS and Brookings definitions are defined using the much richer US occupation classification schemes - these have been mapped as closely as possible to the equivalent UK SOC2010 classification scheme. As this is neither an exact nor standardised process - other analysts may come up with different mappings.

2. APPENDIX OF SIC TABLES

TABLE A2.1 SIC07 CREATIVE INDUSTRIES AND THEIR SEGMENTS

Creative services industries

TABLE A2.2 SIC07 HIGH-TECH INDUSTRIES

TABLE A2.3 HIGH-TECH AND CREATIVE INDUSTRY CODES COMPARED (SICO7)

3. APPENDIX ON EMPLOYMENT INTENSITY DISTRIBUTIONS

TABLE A3.1 INTENSITY DISTRIBUTION OF EMPLOYMENT CORRESPONDING TO THE OCCUPATION CLASSIFICATIONS CONSIDERED IN THIS REPORT (AVERAGE OF 2011 TO 2013)

Note: The figure in the leftmost column gives a range of intensities or share of employment within industries. The figures in the remaining columns specify how much STEM employment is to be found in industries whose intensity falls in the given range of frequencies.

• = sample too small to disclose

4. APPENDIX ON USING INTENSITY AS A DISCRIMINATOR BETWEEN INDUSTRIES

When intensity (or any other continuous indicator) is used to separate industries into two groups A and B, it is implicitly assumed that the industries in each group form a distinct statistical distribution. The intensity of any given industry is then a function of the group to which it belongs, together with other factors which are not entirely known. If the two groups have widely-differing means, and small standard deviations compared to the differences between the two means, and if both parameters are known, this offers an intuitive method of determining which group any given industry belongs to. Suppose that group A has a low average intensity, μ(A) and group B a high average intensity, μ(B). A 'threshold' frequency f* can then be set, with μ(A) < f* < μ(B) such that any industry with an intensity less than f* will be assigned to group A, and any industry with an intensity greater than f* will be assigned to group B.

Difficulties arise if the distribution of intensities within the two groups is not known and in particular if μ(A) and μ(B) and the associated standard deviations are not known. In the case of the creative industries, the frequency distribution of employment as a function of intensity seems more clearly bimodal, and estimates of the population parameters of the creative and non-creative industries can be established, as was done in Dynamic Mapping by establishing a 'most likely' threshold between the two modes, so that statistical estimates of the population parameters, and the equiprobable separation between them, are established at the same time.

When studying the relation between STEM occupations and high-tech industries, this method cannot be used because the distributions are less clearly bimodal. In this situation, we must make use of additional information to inform industry classifications.

5. APPENDIX ON INDUSTRIES WITH A STEM INTENSITY OF MORE THAN 15 PER CENT(WITH MORE THAN 4000 EMPLOYEES) THAT DO NOT FALL WITHIN THE EUROSTAT HIGH-TECH DEFINITION

TABLE A5.1 NON-EHT INDUSTRIES WITH HIGH STEM INTENSITY

ENDNOTES

1 This is on the basis of employment in both first and second jobs. That is, an individual who holds down two separate jobs in the creative industries is counted twice. 2 Chapain, C., Cooke, P., De Propris, L., MacNeill, S. and Mateos-Garcia, J. (2010) 'Creative Clusters and Innovation: Putting Creativity on the Map.' London: NESTA. 3 Higgs, P., Cunningham, S. and Bakhshi, H. (2008) 'Beyond the Creative Industries: Mapping the UK's Creative Economy.' London: NESTA. 4 A concentration being defined here as an area where the proportion of the workforce employed in the creative (high-tech) economy is more than 1.2 times the proportion that the creative (high-tech) economy constitutes of the UK workforce. 5 Brinkley, I. (2006) 'Defining the knowledge economy.' London: The Work Foundation. 6 HM Government (2013) 'Information Economy Strategy'. London: BIS. 7 Bakhshi, H., Freeman, A. and Higgs, P. (2013), 'A Dynamic Mapping of the UK's Creative Industries.' London: Nesta. 8 HM Treasury and BIS (2011) 'The plan for growth.' London: HM Treasury and BIS. p3. 9 Bakhshi, H., Hargreaves, I. and Mateos-Garcia, J. (2013) 'A Manifesto for the Creative Economy.' London: Nesta. Osborne, M.A. and Frey, C.B. (2013) 'The future of employment: How susceptible are jobs to Computerisation?' Oxford: Oxford Martin School, University of Oxford. 10 DCMS (1998) 'Creative Industries Mapping document.' London: DCMS. 11 U.S. Congress, Office of Technology Assessment (1982) 'Technology, Innovation, and Regional Economic Development.' Washington DC: U.S. Government Printing Office. 12 This report (Bakhshi, Freeman, and Higgs 2013) is frequently mentioned, and so for ease of reading we refer to it henceforth as 'Dynamic Mapping'. 13 Hecker, D. (2005) High-technology employment: a NAICS -based update. 'Bureau of Labor Statistics, Monthly Labour Review.' Washington DC: Bureau of Labor Statistics. 14 Since the publication of 'Dynamic Mapping' the DCMS has adopted the methodology in the production of the official DCMS creative industries economic estimates. There are, however, some small differences in the assessment of which occupations are creative, and in the case of libraries and museums, for example, the DCMS has included industrial codes whose creative intensity is below the 'threshold' adopted in the 'Dynamic Mapping'. For consistency with the published official statistics we use the DCMS classifications in this report. See DCMS (2015) for the latest official creative employment estimates. 15 R&D intensity in manufacturing is measured by the ratio of an industry's R&D expenditure to value added, and knowledge intensity in services is measured by its share of tertiary educated persons at NACE 2-digit level. Only services labelled 'high-tech' by Eurostat are considered in our analysis. 16 DCMS (1998 and 2001) 'Creative Industries Mapping Documents.' London: DCMS. 17 Garnham, N. (2005) From cultural to creative industries. 'International Journal of Cultural Policy.' 18 Smith, C. (1998) 'Creative Britain.' London: Faber & Faber. 19 Bakhshi, H., Hargreaves, I. and Mateos-Garcia, J. (2013) 'A Manifesto for the Creative Economy.' London: Nesta. 20 Previous studies that have emphasised the importance of understanding the occupation-industry mix of economies include Currid and Stolarick (2010) The Occupation-Industry Mismatch: New Trajectories for Regional Cluster Analysis and Implications for Economic Development. 'Urban Studies.' 47(2); and Mellander, Stolarick and King (2011) 'What You Do, Not Who You Work For: A Comparison on the Occupational Industry Structures of the US, Canada and Sweden.' Martin Prosperity Institute Working Paper, University of Toronto. 21 DCMS (2014) 'Creative Industries Economic Estimates'. London: DCMS. 22 HM Treasury and BIS (2011) 'The plan for growth.' London: HM Treasury and BIS. p3. 23 Westlake, S. (2013) Rebalancing Act: Rationales and Policies for Sectoral Economic Rebalancing. 'Oxford Review of Economic Policy.' 24 Speech by Rt Hon David Willets MP (2012) 'Our High-tech future'. 25 Government Office for Science (2012) 'Technology and Innovation Futures: UK Growth Opportunities for the 2020s.' London: Government Office for Science. 26 Rt Hon David Willets MP (2013) 'Eight Great Technologies.' Speech at Policy Exchange. 27 Dyson, J. (2010) 'Ingenious Britain. Making the UK the leading High-tech exporter in Europe.' 28 DTI (2006) 'Science, Engineering and Technology skills in the UK.' London: DTI. 29 DFES (2006) 'The Supply And Demand for Science, Technology, Engineering And Mathematics Skills In The UK Economy.' London: DFES. 30 DIUS (2009) 'The Demand for Science, Technology, Engineering and Mathematics (STEM) Skills.' London: DIUS. 31 BIS (2011) 'STEM graduates in non-STEM jobs'. London: BIS. 32 UKCES (2011) 'The Supply of and demand for High-Level STEM skills.' See also UKCES (2013) 'Supply of and demand for High-Level STEM skills.' 33 UKCES (2011), op. cit., p4. 34 UKCES (2013), op. cit. http://www.ukces.org.uk/publications/er77-high-level-stem-skills-supply-and-demand 35 KPMG and Tech Monitor UK (2013) 'Understanding tech clusters and tracking the UK tech sector's outlook for employment and growth.' 36 BIS (2012) 'BIS industrial strategy: UK Sector analysis.' London: BIS. 37 BIS (2012) 'BIS industrial strategy: UK Sector analysis.' p9, Box 2. London: BIS. 38 Bakhshi, H., Freeman, A. and Higgs, P. (2013) 'A Dynamic Mapping of the UK's Creative Industries.' London: Nesta. 39 See ONS (2010) 'Standard Occupational Classification 2010.' Newport: ONS. 40 These are whether the occupation: 1. Involves a novel process, 2. Is mechanisation resistant 3. Is non-repetitive or performs a non-uniform function 4. Involves a 'creative' contribution to the value chain irrespective of context and 5. Involves interpretation, and not mere transformation. Different occupational codes at the four-digit level are assessed against these criteria, based on the detailed list of sub-occupations (the 'coding index') published by the ONS (ONS, 2010). 41 In a forthcoming report we use machine learning-based techniques to identify 'creative occupations' from the official codes. Almost all of the occupations scored as creative in 'Dynamic Mapping' also turn out to be creative using this approach, although a relatively large number of occupations not classified as creative in 'Dynamic Mapping' also turn out to be creative. 42 Hecker, D. (2005) High-technology employment: a NAICS-based update. 'Monthly Labour Review.' U.S. Bureau of Labour Statistics. 43 Bay Area Council Economic Institute Report (2012) 'Technology Works: High-Tech Employment and Wages in the United States.' San Francisco CA: Bay Area Council Economic Institute. 44 Echeverri-Carroll, E. and Ayala, S. (2009) Wage differentials and the spatial concentration of high-technology industries. 'Papers in Regional Science.' Vol. 88, Issue 3, pp.623-641. Note other studies in this area include Markusen, A., Chapple, K., Schrock, G., Yamamoto, D. and Yu, P. (2001) 'Gauging Metropolitan "High-Tech" and "I-Tech" Activity.' which uses a different threshold of three times the national average. 45 Rothwell, J. (2013) 'The hidden STEM economy.' Washington DC: Brooking Institution Metropolitan Program. The O*Net data is part of an ongoing project funded by the Department of Labor's Employment and Training Administration to provide detailed information about every occupation in the US economy. 46 We had originally hoped that the analysis could be expanded to include data from the 2011 Household Census as this also provides occupational and industry data and, in principle, covers the whole population so is not subject to sampling biases. In practice, however, the UK Census has no benefit for our research over the APS/LFS. First, because the Census utilises only 200 industry classifications at a mixture of 2-, 3- and 4-digit level of detail out of the SICO7 standard of 617 codes available at 4 digits of which 603 are used in the APS. In the Census only 21 4-digit industry codes used, and mostly relate to public and government activities. As a result, the ability to discriminate creative from other industries is significantly reduced from 30 industries under the APS to just 15 industries in the Census. Second, we cannot undertake regional analysis on census data as there appears to be no 4-digit occupation within the 3-digit industry employment tables released at the regional or sub-regional level. And third, because the Census is conducted only once every ten years it is unsuitable for time series analysis for anything other than long-term trends. 47 ONS (2011) 'Labour Force Survey. User Guide, Volume 1.' p7. Newport: ONS. 48 DCMS (2014a) 'Creative Industries Economic Estimates.' London: DCMS. 49 Note that these differ from those published in DCMS (2014a) which are based on 2011 and 2012 averages. 50 For consistency with the published DCMS statistics, all estimates of creative and high-tech employment in this report include the total of main and second jobs. 51 The equivalent table in the previous 'Dynamic Mapping' report is Table 10.2. 52 DCMS (2014a)'Creative Industries Economics Estimates, January.' London: DCMS. 53 The rationale for segmentation, given in Cunningham (2014) arises from work on Australian Census data at the most detailed level of industry and occupation classifications that are possible. CCI found correlations in their characteristics sufficient to allow them to be gathered into these two distinct sub-categories. In the UK context a similar distinction was also made in Nesta's (2006) 'Creating Growth' report, Appendix 3 - A refined model of the creative industries pp53-55. 54 See also Bakhshi, H., Freeman, A. and Higgs, P. (2013) 'A Dynamic Mapping of the UK's Creative Industries.' London: Nesta. Table 2.4, p13. 55 Ibid, p54. 56 The geography analysed in this sub-section is the European NUTS1 geography, which corresponds to Government Office Regions (GORs) in England and to the devolved administrations elsewhere in the UK. Location is measured on place of residence as opposed to place of work, so in some cases people may be working in one region and living in another, for example London and the South East of England. For more information see: http://www.ons.gov.uk/ons/guide-method/geography/beginner-s-guide/eurostat/relationship-of-nuts-to-uk-administrative-geographies.html 57 Freeman, A. (2009) 'London's creative workforce: 2009 update.' London: Greater London Authority. 58 Freeman, A. (2009) uses the term microspatial clustering. This work, based on IDBR data, calculated an index of geographical concentration at the level of the ONS's Medium Super-Output areas - approximately the size of electoral wards. Chapain et al., (2010) describe the 'dartboard effect' of microspatial clusters peppered across the UK that are hidden in the analysis of agglomeration economies at higher levels of aggregation. 59 The combined population of Manchester and Merseyside was 4.6 million according to mid-year estimates released by ONS on 17 May 2011. The population of the West Midlands was 5.6 million on March 27 2011, census day (http://www.ons.gov.uk/ons/rel/mro/news-release/census-shows-increase-in-population-of-the-west-midlands/censuswestmidlandnr0712.html). The population living within the Greater London Authority area was 7,825,200 on census day. A better comparison may be afforded by the population of the Greater London Functional Urban Region (Freeman and Cheshire 2006:5) of which the best available estimate is 12,255,000. This is still less than three times the population of Manchester and Merseyside and barely double that of the West Midlands. 60 For a discussion of the role of culture in world cities see BOP (2014) 'World Cities Culture Report'. London: BOP Consulting. 61 Hecker, D. (2005) High-technology employment: a NAICS -based update. 'Bureau of Labor Statistics, Monthly Labour Review.' Washington DC: Bureau of Labor Statistics. 62 DIUS (2009) 'The Demand for Science, Technology, Engineering and Mathematics (STEM) Skills'. London: DIUS. 63 BLS (2010) 'Options for defining STEM (Science, Technology, Engineering, and Mathematics) occupations under the 2010 Standard Occupational Classification (SOC) system: SOC Policy Committee recommendation to the Office of Management and Budget (OMB).' Washington DC: Bureau of Labor Statistics. 64 Greenwood, C., Harrison, M. and Vignoles, A. (2011) 'The labour market value of STEM qualifications and occupations.' Report prepared by the Institute of Education for the Royal Academy of Engineering. 65 Rothwell, J. (2013) 'The hidden STEM economy.' Washington DC: Brookings Institution Metropolitan Program. 66 These codes were mapped to the UK equivalent using a combination of techniques. Approximately half of the occupation codes were able to be mapped using a two-step process using the US SOC to ISCO 08 (International Standard Classification of Occupations 2008) cross walk as the first step and then using the ONS-supplied ISCO to UK SOC2010 crosswalk file as the second step. The balance of occupation codes were mapped using a manual process of matching the main job description for each occupation and identifying the most equivalent occupation title in the ONS's coding index of UK occupations. 67 Eurostat 'Aggregations of manufacturing and services based on NACE Rev 1.1.' 68 High-tech in manufacturing is defined using the intensity of expenditure in R&D and knowledge intensity is defined using the industries' employment share of tertiary educated people. It should be noted that most are 2-digit codes, as a result of which all the underlying 4-digit codes are included in our initial Eurostat High-Tech list. 69 OECD Directorate for Science, Technology and Industry (2011) 'ISIC rev.3 Technology intensity definition.' Paris: OECD. 70 Each 2-digit or 3-digit NACE industry was expanded to the SICO3 industry at the 4-digit level of detail. The UK industry coding index was then used to map the SIC03 industry activities to the most relevant SIC07 industry. 71 Industries with intensities above 15 per cent that were not classified within the Eurostat definition and have over 4,000 employees are included in Appendix 5. 72 In order to be included the industry had, in addition to passing the intensity test, had to have a workforce sample size that was greater than 4,000 for the average of 2011-2013. 73 Chapain et al., (2010) 'Creative clusters and innovation.' NESTA. 74 Helpman, E. and Trajtenberg, M. (1996) 'Diffusion of General Purpose Technologies.' Cambridge MA: NBER Working Paper No. 5773; Rincon, Vecchi and Venturini (2013) 'ICT as a General Purpose Technology: Spillovers, Absorptive Capacity and Productivity Performance.' NIESR discussion paper 416. London: NIESR. 75 In ongoing research we are exploring the use workforce intensity of ICT occupations to identify Information Economy industries. 76 e-skills UK, Intellect and BCS (2013) 'Information Economy Economic estimates 2013.' London: e-skills UK, Intellect and BCS. 77 In other research we have used data sourced from public online sources, as opposed to official survey data, to map the location of the video games industry at the hyperlocal level (Mateos-Garcia, Bakhshi and Lenel, 2014). 78 http://epp.eurostat.ec.europa.eu/portal/page/portal/nuts_nomenclature/introduction 79 In new research we are applying the 'Dynamic Mapping' approach to EU and North American countries. 80 The location quotients are calculated based on employment figures for main and second jobs in UK NUTS3 areas 2003 geographies (mapped using the 2006 geographies available from the Eurostat website). 81 West Inner London is made up of City of London, Camden, Hammersmith and Fulham, Kensington and Chelsea, Wandsworth and Westminster. 82 East Inner London is made up of Hackney, Haringey, Islington, Lambeth, Lewisham, Newham, Southwark and Tower Hamlets. 83 West and North West Outer London is made up of Barnet, Brent, Ealing, Harrow, Hillingdon, Hounslow and Richmond upon Thames. 84 South Outer London is made up of Bromley, Croydon, Kingston upon Thames, Merton and Sutton. 85 Source Rolls Royce website. 86 Source Sellafield website 87 Midlands Aerospace Alliance 'Mobilising the Midlands aerospace cluster.' Coventry: Midlands Aerospace Alliance. 88 Nathan, M. and Vandore, E. (2014) Here Be Startups: Exploring a young digital cluster in Inner East London. Forthcoming in 'Environment and Planning.' 89 Florida, R. (2012) The Joys of Urban Tech. 'The Wall Street Journal.' 31 August 2012. 90 KPMG and Tech Monitor UK (2013) 'Understanding tech clusters and tracking the UK tech sector's outlook for employment and growth.' p9. 91 The range is the difference between the largest and the smallest LQ for an industry grouping across the NUTS3 geographies. Standard deviation is a measure of dispersion which is based on the square of deviations from the mean. The Gini coefficient is a measure of equality among the LQs, which is bounded between O (all LQs are the same) and 1 (Complete inequality, only one NUTS3 area has a positive LQ). 92 Bakhshi, H., Freeman, A. and Higgs, P. (2013) 'A Dynamic Mapping of the UK's Creative Industries.' London: Nesta. 93 Bakhshi, H., Lee, N. and Mateos-Garcia, J. (2013) represents an attempt to do this using earlier classifications for creative occupations and industries.

Nesta...

Nesta 1 Plough Place London EC4A 1DE [email protected] @nesta_uk www.facebook.com/nesta.uk www.nesta.org.uk

Nesta is a registered charity in England and Wales with company number 7706036 and charity number 1144091. Registered as a charity in Scotland number SCO42833. Registered office: 1 Plough Place, London, EC4A 1DE. CC i BY NC SA