Dr Jennie Jordan, CRAIC, Loughborough University London
Creative Horizon 5 is a research collaboration between CRAIC, the Creative Innovation and Research Centre at Loughborough University London, Creative Informatics, Edinburgh’s creative cluster, and The Data City, a platform using machine learning to add value to data on innovation in the UK economy. The collaboration asked what are the data (adoption) barriers to innovation in the creative industries; and how can we make data collection, processing and analysis more useful for data users like policy makers and funders and data providers, such as creative businesses and individuals? You can read more about the research questions in our earlier essay.
The research has to date gone through three phases. Firstly, we undertook in-depth interviews with 28 individuals who either provide, collate or use data within the creative industries. We asked them how they defined data and innovation, how they engaged with and used data, and what problems they associated with data in the sector. Secondly, we ran four interactive workshops with creatives and policy makers, introducing them to The Data City’s platform as an example of new tools available to aggregate and analyse data using machine learning. We also held a ‘platforms’ roundtable to identify common issues and potential for collaboration between a number of data platforms working in the sector. Finally, we ran a policy hack day at the end of September asking policy makers and trade bodies ‘what actions can be taken to fill data gaps, make data more transparent and drive innovation in creative industry ecosystems’ and, of those, ‘what can they take forward in their organisations?’
Creative ecosystems can be conceptualised as having six core dimensions – strategy/policy; tangible and intangible infrastructure; funding/investment; education (learning and skills); communities; and content creators. The first three are place based, but with intersections between scales. Creatives might be affected by local, regional, national or international policies and investment decisions. In relation to innovation and R&D, copyright and trademark regulations are supranational, for example. The final three relate to people and communities. Who are the content creators, what do they know, and how is this knowledge circulated and developed? What are the demographics of the ecosystem, and how do interconnections work? How open or closed are its networks?
Dark matter: what is R&D in creative ecosystems?
Effective support to grow innovation ecosystems requires, first, a good understanding of each of these elements and how they interact. Ecosystem mapping is a relatively new area of work within the creative industries and data collection is uneven. And gaps can be substantial. In September 2022 the Office for National Statistics revised its estimate of R&D spend from 1.7 per cent of GDP to 2.4 per cent. The adjustment was instigated by differences in the numbers of R&D tax credit claims, and the estimates of spend drawn from the ONS’ annual Business Enterprise Research and Development (BERD) survey. One criticism of this survey is that its methodology is flawed because of a belief that most businesses do not undertake R&D, and that those that do are likely to be larger firms. This is a crucial insight for the creative industries where 94 per cent of registered companies are micro-businesses employing fewer than 10 staff.
A second criticism of government statistics in relation to innovation and R&D relates to definitions. The international standard is the OECD’s Frascati Manual. This defines activities as innovative if they fulfil all of the following: novel, creative, uncertain, systematic, and transferable. This fails to capture much of the innovation in, for example, software and gaming, or the spillovers from CreaTech into non-creative industries, such as the use of animation and 3D modelling in textile sample production or bridge building. But as Andrew Chitty argues even this definition is more inclusive of creative industry innovation than that used by the UK Government for its R&D tax credit which requires R&D to be for scientific and technological advancement, and specifically excludes arts, humanities and social sciences. So creative innovation, such as a new TV format, would be excluded.
If these intangible assets, known as Intellectual Property Products (IPPs), which include entertainment, literary and artistic originals and software, are incorporated into data, the UK’s R&D spend is much closer to the OECD average of 2.4 per cent of GDP. When other assets that are features of the creative industries, such as branding and design processes and products, are incorporated, the UK overtakes the OECD average. These conflicting analyses illustrate the dominance of concepts of R&D based in a particular sort of industrialised manufacturing economy, an economic model that may never have been entirely accurate, but which is now being challenged by synergies between the knowledge and creative economies and spillovers from these into other sectors.
Dark matter: know-how and know-who
A third criticism of the statistical models used for understanding innovation is the extent to which they overlook know-how, know-who and knowledge exchange. Our research highlighted the importance of iterative processes of skill development among creatives, the ‘doing, using and interacting’ (DUI) mode of innovation, a process described by one research participant as “analogue data”. Interviewees talked about their process in terms that reflected the DUI model factors of learning by doing to develop skills and know-how; learning by using – a key element in design thinking and user experience (UX) approaches to innovation; and know-who, learning by interacting with other companies in and out of your sector, with suppliers and competitors and informally with former colleagues. Interaction is driven in the creative industries by individuals moving from company to company, project to project, building expertise, and sharing that in their new roles, and by knowledge-exchange facilitators and hubs such as artist studios, technicians in specialist workshops, festivals and university art and design courses. Methods of capturing innovation as process, as knowledge flow, rather than simply through inputs and outputs are widely discussed in the innovation literature, with a number of proposals posited for capturing the dynamism of knowledge and creative innovation.
Dark matter: what is measured counts
Despite these issues, and their policy implications having been extensively debated, what was evident from policy makers in our research is that what is measured counts more than narrative or case study research. There is a bias towards statistical data and studies undertaken first by the Office for National Statistics that provide international comparability, then by apparently impartial organisations, for instance universities or international bodies such as UNESCO, which is generally believed to be more robust than case studies. Where such data is not available, trade bodies provide valuable insights on the experiences of their members. Their research is taken seriously by policy makers, but is seen as having a vested interest, so methodologies are rigorously scrutinised. However, even when such data was found to be robust, there was a lack of comparability across studies and subsectors, few longitudinal studies that could track change, and limited focus on “how different sectors are linked and the exchange of information between them”.
On a positive note, it was felt during the research that “recognition of the invisibility of sole trader creatives is growing”, and there was a desire to know how “to navigate ‘dense’ data networks to find what connects me to other creatives” outside existing industry niches. There was great interest in the potential of The Data City’s model for finding and illustrating interconnections within and outside the ecosystem, albeit with caveats about the quality of the data being used to generate it. Other data platforms were aggregating data on audiences, on events and markets. Where they shared elements of their data in open source and standardised ways, they were finding that it was enabling new products and services to be developed, and they expressed a desire to “work harder/better towards standardising approaches (even across competitors)”.
Data matter = questionable analysis and inferior policy support
Three highly distinctive features of creative industry ecosystems have emerged as central to innovation, and potentially quantifiable, but with no accessible systematic data collection mechanism.
- Freelancers comprise at least 32 per cent of the sector’s workforce and 76 per cent of creative industries companies had worked with a freelancer over the 12 months to March 2020. In what is a project and microbusiness-based ecosystem, freelancers were described by one interviewee as “pockets of knowledge” moving between organisations and, potentially, catalysing innovative spillovers as a result. What (who) are the linkage points between different sectors where information is exchanged? The Unique Taxpayer Reference (UTR) number given to anyone who is required to fill in self-assessment tax returns was identified as a potential common data standard for collating information on sole traders, but this is not currently open source in the way company numbers are.
- Much innovation in the creative industries is “analogue”, based on DUI-learning; know-how, know-who, expertise and judgement. While there are methods posited for collecting and aggregating such data, few are utilising new technologies such as the semantic web scraping and machine learning that The Data City has developed to enrich basic innovation data.
- And copyright, the intellectual property regulation most relevant to value creation in the sector, is not registered in the same way that patents are, meaning many creative innovations are not included in measures such as the World Intellectual Property Organisation’s (WIPO) Global Innovation Index (World Intellectual Property Organisation (WIPO), 2022). As with freelancer data, there are potential sources here such as the collective management organisations that collect and distribute royalties to creatives.
Until light is thrown on this dark matter, any statistical analysis of creative industries ecosystems is questionable, undermining policy and programme development. Finding systematic, real time open data on these elements of the ecosystem would greatly enhance policy-making in the sector since freelancers are significant creators of innovation, and copyright the primary mechanism for capturing value in the creative industries, respectively.
We will be further interrogating our data over the next few months with the aim of publishing a white paper. In the meantime, if you would like to know more, please do get in touch with Jennie Jordan (j.jordan@lboro.ac.uk) or Graham Hitchen (g.hitchen@lboro.ac.uk).