Digital Automation Technologies and the Future of Work(ers): A Policy Roadmap
Maria Savona (University of Sussex, SPRU) and Brian MacAulay (Digital Catapult)
With the accelerating pace of recent years, there is an urgent need for policymakers to address the current and future impact of emerging digital automation technologies on jobs, skills demand, and wage inequality. Critical decisions will need to be made to ensure that investment, education, and skills training keep up with a range of technological and socioeconomic shifts. Effective policies for a just transformation should be grounded in both a knowledge of past waves of Information and Communication Technologies (ICTs) on labour markets as well as ongoing research.
The H2020 PILLARS Project is at the forefront of current studies of present and future shifts in jobs and skills demand in labour markets driven by changes in technology, trade, and industrial dynamics. This Policy Brief summarises key project findings alongside policy recommendations generated at a multi-stakeholder Policy Workshop held at Digital Catapult London on the 5th December 2023.
Emerging Digital Technologies (EDT): Exposure and Adoption
The PILLARS project analysed a large corpus comprising millions of patents and scientific publications related to automation technologies, and identified novel and emerging digital technologies. The research combined intelligence from hundreds of experts, and made predictions about which automation technologies will become prevalent by 2030, which tasks they will perform, and what skills they will need. The researchers employed state-of-theart sentence transformers from computational linguistics, and estimated which occupations and industries are most exposed to the tasks performed by emerging automation technologies, and how this has impacted on regional labour markets.
Finally, the PILLARS team examined the extent to which industries open vacancies to hire workers, in the UK and other European countries.
Key Findings:
- Besides generative large language models, the most novel and fast growing scientific and technological development are led by cloud computing, secure data infrastructures such as blockchain, advanced additive manufacturing, and mobile and collaborative robots.
- Non-routine cognitive tasks such as those of clerical support workers, technicians, professionals, and managers are becoming highly exposed to emerging digital automation technologies, with managers positioned in the middle of the exposure distribution.
- While firm adoption of most automation technologies is still low, the most exposed regions in the UK and other European countries have experienced an increase in employment, except regions that are exposed to embedded systems such as industrial automation, the Internet of Things and remote monitoring.
- UK firms are lagging behind in the adoption of most emerging technologies, including those in which service-oriented countries are leading, such as machine learning, and cloud computing, and even more so in industrial automation.
Technology Investments and Regional Labour Markets
PILLARS has investigated the long-term evolution of EU and UK regional investments in digital automation technologies over the life cycle of ICT and Robot developments between 1995-2017. Major technological breakthroughs during this period have been identified and associated with phases of acceleration and deceleration in investment.
Long-term and short-term exposure to automation technology affects employment and wages differently. In the short-term, for the different phases of acceleration and deceleration of each breakthrough, PILLARS research found that the negligible long-term impact of automation on employment conceals significant short-term positive and negative effects within phases of the technology life cycle. The research also found that the negative impact of ICT investments on employment is driven by the phase of the cycle when investment decelerates (and the technology is more mature). The phases of the technology life cycles are more relevant than differences in regions’ structural characteristics, such as productivity and sector specialisation in explaining the impact of automation on regional employment.
Skills and On-the-Job Training
PILLARS has employed machine learning techniques that link data on skills demand and supply and developed new measures to investigate two key aspects: (1) the prevalence of skills mismatches across occupations and regions in the UK, and (2) how the degree of skills mismatch in the UK compares to other European countries. The results show that on-the-job training can be an effective measure to upgrade workers’ skills and be ready for automation. However, the demand for and efficacy of training might depend on labour market tightness and the level of employer protection. PILLARS' quantification of skills mismatches across occupations and regions serves as a basis for discussions on the effectiveness of training for both firms and workers to strategically adapt to the evolving requirements of the labour market.
Recommendations for Policy Makers and Firms
1 . Foresight Techniques
Develop and adopt foresight techniques for government, firms and workers to prepare and adapt to emerging technologies, and to unlock their potential, rather than adapt once they have been introduced.
2 . Monitor Evolving Technologies
Monitor the evolution of technologies and the overall level of (un)employment at certain critical points when radical breakthroughs, e.g. Generative AI, create both new emerging tasks and the transition from old to new tasks.
3 . Invest in Research
Support firms to invest in research on EDTs, rather than only adoption, so that they can catch up with the cycles of technology and investment that shape the evolution of technological breakthrough.
4 . Release Untapped Potential
Release the untapped potential of firms and regions, especially those lagging in productivity, through productivity improvements associated with adoption of EDTs.
5. Inclusive Governance
Ensure inclusive governance of data with the increasing use of data-intensive EDTs in individual activities, digital trade, AI research and other spheres.
6 . Intellectual Property
Address the boundaries between human and artificial intellectual property rights from the perspective of both antitrust law and value redistribution.
7. Training
On-the-job and vocational training in EDTs delivers most value when combined with other complementary management and soft skills.
8 . Citizens and Workers' Voices
Citizens and workers’ (e.g. platform workers) voices must be represented in decisions to support critical technologies where the effects on working conditions may worsen.
References
Chaturvedi, S., Prytkova, E., Ciarli, T., Nomaler, Ö, 2024, What is the Future of Automation? Using Semantic Analysis to Identify Emerging Technologies, PILLARS Report.
Giabelli, A., Prytkova, E., Petit, F., Ciarli, T., 2024, Advertised Technologies: Identifying Adoption of Emerging Technologies in Online Job Postings, PILLARS (mimeo).
Oliveira, A., Ciarli, T., Consentino de la Vega, R., Scholten, C., 2023, The Future of digital automation technologies tasks and skills, A Delphi survey, PILLARS (mimeo).
Prytkova, Petit, Li, Chaturvedi, Ciarli, 2024, The Employment Impact of Emerging Digital Technologies, PILLARS Deliverable 3.1.
Jaccoud, Ciarli, Petit, Savona (2024), Automation and employment over the technology life cycle: Evidence from European regions, Pillars Deliverable 1.2.