Major milestone for P4W: national skills survey complete

Developing insights in supply and demand on the labour market is one of the core activities of Passport4Work. This is done by painting a timely picture of the skills needed by employers, and those in possession by workers, and then comparing the two.

One important prerequisite for this comparison is the realization of a common understanding of skills. After all, skills can be complex and abstract, leaving plenty of room for interpretation. To avoid this, Passport4Work made the decision to build its tools on a universally known skills language (also known as a taxonomy): the US-based O*NET. However, a major concern has been the representativeness of O*NET for the Dutch labour market. Considering the highly contextual nature of occupations and skills, merely copying the American skills language and its underlying data could result in mismatches due to a misrepresentation of the Dutch labour market. This could occur because O*NET includes information on the relevance, and required level of a skill for a certain occupation. 

In an attempt to still benefit from the recognizability and robustness of the O*NET skills language, but to make sure the underlying data on the relevance and required level of a skill are accurate for the Dutch labour market, a national survey was conducted among Dutch employers and employees (which is the same methodology employed by O*NET in the United States). 

The survey was built around the three sectors of industry which are central to Passport4Work: Health care, construction, and the technical industry. For each of these sectors, 8 key occupations were selected together with industry representatives (looking at criteria such as demand, and foreseen importance for the near future).

While the survey was conducted in the middle of the pandemic, the response has been satisfactory. In total, 1139 employers and employees contributed to the survey. The confidence level of the results is 95%, with a margin of error of 8% for construction, 7% for health care, and 10% for the technical industry (due to slight differences in the response rate across the sectors). This means that if we were to repeat the survey 100 times, more or less the same results would be obtained in 95 cases.

While the project  team is still analyzing all the results, a few preliminary findings are worth sharing:

• In the survey we also asked for the most important 21st century skills. The top 3 is composed of communication, problem solving, and working in a team. Somewhat surprisingly, ICT skills rank in the bottom half of the top 10. This finding is particularly interesting in the context of the robotics and automation debate; the top-3 reflects those skills which are most difficult  to automate.
• On first glance, all occupations have completely different top 10's in terms of the most relevant skills. Across sectors, problem solving, communication and listening are recurring skills in the upper echelons of the data.

The results will be used in three ways. First of all, it will be shared with the national skills language initiative CompetentNL (initiated and governed on a national level). By doing so, P4W provides evidence-based information on the relevance and required mastery level of skills. On its basis, it could be upscaled further to encompass all sectors of industry in The Netherlands. 

Second, the data will drive the matching algorithm in Passport4Work. In essence, this means that a jobseeker's completed skills passport (in P4W terms, their "skillprint) will be matched with a particular vacancy based on this data.

Third and finally, the data is included in P4W's vacancy builder tool. Here, employers can simply click on a occupation for which they have an open vacancy, and the survey data is pre-filled for them in the system. This reduces a big burden for employers, who no longer have to analyze which skills are most important for a occupation.