As part of the Digital Futures of Work Programme, we brought together over 30 experts from Europe and Asia to discuss new methods to improve our understanding of skills anticipation, job redesign, and labour markets in a context of rapid digital innovation. There is increasing interest in the potential and limitations of labour market information (LMI) and these are some of the learning points from that discussion.

The new does not displace the old: the need for multiple data sources

A central theme in many of the discussions revolved around the need to combine multiple sources of data (including company and employee interviews and surveys, job vacancy data, investment data, data on patents, education providers data, amongst others), moving us on from discussions on what data sources should be used, to better understand how to integrate different sources. For example, employer surveys tend to offer broader coverage of issues, including work systems and business strategies, but also less granularity on skills demands than big data analysis from job postings, which can also be analysed in real-time. In short, there is no perfect method, and which to select depends on the aim of the analysis.

The need for multiple voices 

Much labour market information is focused on the demands and perceived needs of employers. But we also need to explore the views of other stakeholders, including employees if they are to be engaged in their work.  A central question raised in the Digital Futures of Work roundtable was how to enhance our understanding of the employee voice, together with those looking for work. What are their views on issues like skills utilization, job quality, career trajectories, wage expectations, motivations, and aspirations for the future of their work?

Need for clarity and consistency

There is a multitude of conceptions and classifications of skills currently in operation. There are also differences in vocabulary: between educational institutions, employers, and analysts. This is confusing for users of labour market data, and also makes comparative and trend analysis more difficult. There is a need to develop common tools and skills taxonomies to advance our understanding of the labour market.

Keep users in mind

There is increasing demand for labour market intelligence, but there are also many different users of this ‘intelligence’: e.g., governments, companies, individual job seekers and learners, training providers, and educational institutions. These users may have different data needs when it comes to the labour market but also different degrees of expertise in interpreting the data. What they have in common is the need for actionable analysis, which can feed into decision-making (e.g., government policy strategy, action plans, educational investment, or career decisions for individuals).

For example, the Singapore government has been placing substantial emphasis on the analysis of emerging and priority skills for the ‘future economy’, including Industry 4.0, the care economy, the green economy, and the digital economy.

Other challenges ahead:

There are a number of other challenges ahead when thinking about how to use labour market information to capture the future of work:

• There is still a paucity of publicly available data on labour market demands that is timely, granular, and user-friendly, as the bulk of data of that sort has been commodified and is not generally shared. Initiatives like CEDEFOP’s Skills-OVATE or NESTA’s Open Jobs Observatory aim to address this situation.
• LMI is biased towards the formal labour market (advertised vacancies) and within this, the upper end, although the representation of other segments are improving.
• Coverage also differs substantially between countries. Middle- and low-income countries tend to experience more acute data quality issues.
• There is still little data on individual trajectories in the labour market, including transitions between occupations.
• Legal and administrative barriers, notably around privacy issues, have not been fully worked out in attempts to integrate different sources of data.
• Lack of ownership of the ‘problem’ (different views on whether ‘labour market’ related problems should be tackled by governments, employers, individuals, or all, and under what institutional structures).

New sources of data have enabled previously unthinkable ways to look at the labour market, and new ways of exploiting these data are emerging. They represent fundamental developments towards a better understanding of the labour market and the continuous monitoring and re-evaluation of our individual and collective actions.