WIREDGOV

This POSTnote looks at specialist data skills in the UK, including for artificial intelligence. It considers demand and supply, workforce demographics, challenges, and initiatives to increase supply.

Documents to download

• ‘POSTnote 697 Data science skills in the UK workforce’ report (527 KB , PDF)

DOI: https://doi.org/10.58248/PN697

Data science is an umbrella term for fields that use large quantities of data to discover actionable insights. Data science projects may involve a team with a range of roles such as data scientists, data engineers, data analysts, and data stewards. These roles require both technical and non-technical skills at a range of levels.

This POSTnote primarily focuses on the skills needed for data analysis and modelling, which require technical skills, such as data visualisation and programming; and non-technical skills, such as communication, creative thinking, and data ethics.

In its 2020 National Data Strategy, the UK Government recognised ‘data skills’ as one of four pillars needed to ensure that the UK benefits from data. Additionally, the 2021 National Artificial Intelligence (AI) Strategy recognised ‘skills and talent’ as core to the UK’s success in AI.

The demand for specialist data skills is growing. In 2023, the World Economic Forum surveyed 803 global companies and found that ‘AI and Machine Learning Specialists’ and ‘Data Analysts and Scientists’ roles were in the top 10 jobs expected to grow fastest between 2023 and 2027.

There is a mis-match between the supply and demand of specialist data skills. A 2020 Ipsos Mori survey of 118 UK public and private sector organisations using AI, or developing AI-led products or services, found that 62% of respondents could not meet their goals because job applicants and existing staff lacked the skills needed to work with AI.

The UK Government has launched several initiatives to develop specialist data skills across the UK. This includes £117 million to train PhD students at AI Centres for Doctoral Training from 2024/25. The Lords Science and Technology Committee has raised concerns that there is a mismatch between the scale of the UK’s STEM (Science, Technology, Engineering and Mathematics) skills gap and the solutions posed by the Government.

Key points:

• Collecting and analysing data offers potential economic and social benefits. Analysis by the McKinsey Global Institute estimated that, by 2030, UK GDP could increase by up to 22% as a result of AI.
• Potential societal benefits could range from climate change mitigation, to improving early detection and diagnosis of cancers by using AI to identify patterns from imaging (MRI) scans that are not readily detected by humans.
• Evidence suggests that the availability of people with specialist data skills in the UK is not sufficient to meet demand.
• A 2021 study estimated that the supply of data scientists from UK universities was unlikely to exceed 10,000 per year, yet there were potentially at least 178,000 data specialist roles vacant in the UK.
• Research finds that certain groups (such as women, those from minority ethnic backgrounds and people with disabilities) are underrepresented in the data workforce. A lack of workforce diversity has the potential to amplify existing inequalities and prejudices.
• Initiatives to increase the number of people with data skills include degree conversion courses, doctoral training centres for PhD students, online up-skilling platforms, apprenticeships, and visas to attract international talent.
• Efforts to reduce the skills gap can be hindered by the inconsistent definition of data skills, organisational culture, the availability of specialist primary and secondary school teachers, and barriers to people moving between sectors.
• A 2022 inquiry by the Lords Science and Technology Committee concluded that a mismatch exists between the scale of the UK’s STEM skills gap and the solutions proposed by the UK Government, “especially given the UK’s ambition to be a science and technology superpower”. It described the Government’s policies as “inadequate and piecemeal”.

Acknowledgements:

POSTnotes are based on literature reviews and interviews with a range of stakeholders and are externally peer reviewed. POST would like to thank interviewees and peer reviewers for kindly giving up their time during the preparation of this briefing, including:

• Members of the POST Board*
• Office for Artificial Intelligence*
• Department for Education*
• Office for National Statistics
• Department for Science, Innovation and Technology
• Dr Caroline Chibelushi, Innovate UK KTN*
• David Bowkett, ESRC UKRI
• Trias Gkikopoulos, Innovate UK
• Sam McGregor, AHRC UKRI
• James Dracott, EPSRC UKRI
• Nigel Armstead, SAS Institute*
• Glyn Townsend, SAS Institute
• Mark Donnelly, BAE Systems
• Richard Hamer, BAE Systems
• Nimmi Patel, techUK*
• Jonathan Hobson, Perspective Economics*
• Dr Christopher Hallsworth, Imperial College London
• Dr Michelle Sahai, University of Roehampton
• Rob Slane, Lightcast*
• Layla O’Kane, Lightcast*
• Dr Becky Allen, University of Sunderland
• Dr Federico Botta, University of Exeter*
• Professor Rob Procter, University of Warwick and The Alan Turing Institute for Data Science and AI*
• Victoria Wade, University of London Careers Service
• Professor Neville Davies, University of Plymouth
• Institute for People-Centred AI, University of Surrey
• Dr Charlotte George, X-NET
• Professor Chris Ponting, X-NET
• Professor David Sims, X-NET
• Dr Cristina Martín, X-NET
• Professor Andrew Blake, Independent Consultant*
• Dr Matthew Forshaw, The Alan Turing Institute for Data Science and AI*
• Andrew Strait, Ada Lovelace Institute
• National Centre for AI in Tertiary Education, Jisc
• Neil Sheldon, Teaching Statistics Trust

*denotes people and organisations who acted as external reviewers for the briefing.