Increased adoption of digital and green technologies pressures workers to adapt their skillsets in line with the changing demand for skills. To support them, the European Union actively promotes lifelong-learning initiatives such as individual learning accounts (ILAs) and micro-credentials ¹ See Council Recommendation of 16 June 2022 on individual learning accounts, 2022/C 243/03, . . The EU recommends that member countries should prioritise training programmes related to green and digital technologies, traditionally associated with STEM (science, technology, engineering and mathematics) education. The EU also recognises the gender imbalance in these fields.

ILAs are seen as a potential tool to mitigate the adverse labour market impact of digital and green transitions, and to reduce structural skill gaps between men and women (European Commission, 2019) ² See also Sofia Fernandes, ‘Making the most of the European Year of Skills’, Social Europe, 6 March 2023, . . Predominantly financed by employers, ILAs enable individuals to accumulate training credits over a given period of time typically conditional on either being employed or actively looking for a job. These virtual training credits, stored in a user account, can then be utilized to cover the cost of participating in the training programmes chosen by the individual.

There are several examples where governments allocate individual oriented financial training measures towards priority areas. While many of these aids typically centre on vocational training paths, like Austria's Bildungskonto or Belgium's Opleidingscheques, certain countries opt for a more targeted approach to address specific labour market requirements. Germany's Bildungsscheck, for instance, emphasises language and ICT training, in addition to basic and vocational education. Similarly, Portugal's Cheque formação funds training programmes aligned with the priority areas set annually by the Institute of Employment and Training (IEFP). ILAs also have the potential to guide participants towards courses in high-demand areas that align strategically with EU objectives, in areas like renewable energy, recycling and digital technologies. For example, the Welsh government allocated £3 million (about €3.5 million), or 6 percent of the budget of the whole ILA programme, to train individuals in ICT and “net-zero skills” ³ See Welsh Government press release of 17 October 2022, ‘£3m skills boost to the digital and green sector’, . .

To examine whether ILAs can effectively align skills development with strategic priorities, we assessed one of the earliest ILA implementations in the EU. France’s Compte Personnel de Formation (CPF), introduced in 2015, is one of the best examples of ILAs (OECD, 2019). Using CPF 2021 training data, we analysed how successful CPF was in promoting participation in STEM training in France. While CPF is expected to assist workers in adapting their skills to meet changing labour demands, it lacks an internal mechanism to ensure the training courses taken are be in line with the skills required in the labour market, as it grants complete autonomy to workers in selecting their training programmes.

Our analysis shows that CPF was useful in promoting basic digital skills and specialised software skills, but it fell short of fostering the acquisition of green skills, advanced digital skills (such as programming and artificial intelligence) and other skills that are part of STEM. CPF also reflects gender imbalances that exist in French formal education. Although the most popular training courses are gender-balanced, less popular and more specialised training courses in STEM areas are male-dominated.

Compte Personnel de Formation

To assess the CPF, we used data provided by Caisse des Dépôts, a French public sector investment fund that is responsible for administering the CPF programme ⁴ Groupe Caisse des Dépôts is a Bruegel member. . Overall, nearly 1.5 million trainees are recorded as participating in 435 different training courses. Using the French Minister of Employment and Training’s vocational training classification, we categorised the courses as STEM or non-STEM ⁵ To arrive at our classification of STEM courses, we had to split several categories of the French ministry’s Formacode classification. For example we split ‘ICT, Arts’ into ‘ICT’ and ‘Arts’ based on the descriptions of the content of training courses. . 

In this classification, the following categories of training courses belong to the STEM family: civil engineering, energy, electricity, ICT, industrial production, mechanics, electronics, medicine, science, software, tools, applications, services. Table 1 reports the most common training courses in STEM and non-STEM categories. 

According to our classification, 17.9 percent of all training courses which operated within CPF in 2021 and 15.3 percent of all trainees who used CPF in 2021 were in the STEM category (Figure 1, left panel). STEM-related courses chosen by CPF users focus predominantly on basic digital skills and specialised software, such as database management and computer-aided design software, with 90.3 percent of STEM trainees opting for such courses. Other significant STEM sub-categories include energy and electricity (4.3 percent of STEM trainees) and ICT, which covers advanced digital skills such as programming and AI (4.1 percent). For example, courses in AI attracted only 90 trainees in 2021. Some other components of traditionally understood STEM, such as science and medicine, attracted less than 0.1 percent of STEM trainees.

Figure 1: Share of STEM trainees (left panel) and composition of STEM category (right panel) 

Source: Bruegel.

To investigate how CPF aligns with the green transition, we categorised courses linked to skills relevant to the green energy economy. This category includes courses related to renewable and nuclear energy and environmental sciences, biodiversity and waste management. Despite the relatively broad scope of this category, participation in green training in CPF is markedly modest. Out of 1.5 million trainees, only 1,625 (a mere 0.1 percent) took courses within this domain. For example, training courses in photovoltaic solar energy attracted only 160 trainees. 

Lack of women in technical fields 

The digital and green transition poses a challenge for another reason: the fields in which there is higher demand for digital and green skills are also known to be gender-biased. Many digital and green skills are taught as part of STEM formal education and there is substantial under-representation of women in many STEM fields, across most developed countries (Deloitte, 2022).

Similar to their EU counterparts, French women make up a small share of STEM graduates at all education levels (Figure 2) ⁶ Many academic fields in France are gender-segregated with either very low or high percentages of women. According to the OECD (2021b), French women accounted for 24 percent of graduates in engineering, manufacturing and construction degrees at the tertiary level, and 19 percent in ICT. Conversely, women were a substantial majority of graduates in the education sector (76 percent) and in the health and welfare sector (74 percent). . Vocational secondary education in STEM areas is particularly unpopular with women. In France in 2021, only 13 percent of graduates in those programmes were women, compared to 16 percent in the EU. The percentage of women is higher among STEM graduates at the tertiary level but still far below parity. In 2021, 32 percent of French STEM graduates were women, compared to 33 percent in the EU. 

When it comes to employment, women are underrepresented among scientists and engineers as well, though the gender gap is narrower compared to the disparity observed in education.

Figure 2: Share of women among STEM graduates and employees, 2021

Source: Bruegel based on Eurostat. Note: The secondary level includes vocational upper secondary education and post-secondary non-tertiary vocational education. 

Some ILA programmes are already involved in attempts to promote gender equality. The previously mentioned Welsh Government programme added a “gender budgeting principle” to its ILA programme, according to which the governmental departments responsible for ILA assessments are required to discuss the impact of the programme on gender equality as one of the goals. Although no such mechanisms were involved in the construction of the French ILA, the performance of the programme with regard to gender parity may still be positive. 

Is there a gender gap in STEM-related CPF training courses?

At first glance, CPF STEM training courses seem to be perfectly gender-equal with 50 percent of trainees being women. However, this is driven by training courses in basic digital skills and specialised software, which attracted many trainees and are gender-balanced (Figure 3). Omitting those courses from the calculation results in the percentage of women among STEM trainees falling sharply from 50 percent to 18.9 percent. The bulk of training courses classified as STEM are far below gender parity (Figure 3). This leads to the conclusion that the French ILA scheme replicates trends known from formal education and does not have a gender-balancing effect. Areas known to be male-dominated in French tertiary education, including engineering, construction, ICT, science and technology (OECD, 2021a), tend to be male-dominated in CPF as well. 

Lessons and recommendations

Governments are expected to establish the requisite conditions to equip the workforce with skills essential for navigating the evolving economy. While in some cases ILAs appear well-suited for this purpose, CPF users predominantly opt for courses that offer general and basic skills, such as Microsoft Office. In the area of the digital economy, this signifies a success for the programme, given that 38 percent of the French population still lacks fundamental digital skills ⁷ France: /dataset/technology-adoption-dashboard. , which are also in high demand from employers (Nurski, 2022). However, fundamental digital skills are no longer considered advantageous but rather the standard, as 90 percent of professions now require basic digital skills. Policies should explore ways to diversify the range of training courses taken by workers, fostering expansion into emerging areas with high labour demand. 

France tried to promote participation in training programmes for high-demand skills by providing digital tools, such as Bob Emploi and jobflix ⁸ See  and . . These tools were designed to assist workers in making well-informed decisions about the training courses they choose to pursue. Despite their potential to influence training choices, the data indicates that their impact has been limited. A more effective approach may involve interventions that do not rely on the digital literacy of workers. Enhanced career guidance and improved offline access to information about in-demand skills and available training opportunities could significantly contribute. Additionally, offering credit top-ups as incentives to participation in training courses relevant to sought-after skills could further motivate users to pursue careers in strategic industries.

Finally, efforts should be made to encourage and support women’s participation in CPF STEM courses. Similar to promoting strategic industries, this can be done by providing directed credit top-ups for certain training courses for women. Career guidance, which is available for CPF users, could also play a beneficial role in women’s participation in STEM courses. In particular, attention should be paid to ensuring that career counsellors are well-equipped to advise women willing to learn STEM subjects. Training providers within the CPF programme can also play a vital role in reducing gender disparities by actively promoting and providing equal opportunities for women to access and succeed in STEM-related training courses.

References 

Deloitte (2022) Rethink STE(A)M education: A sustainable future through scientific, tech and humanistic skills, Fondazione Deloitte, available at 

Nurski, L. (2022) ‘Beyond the training gap: learning foundational skills on the job’, Bruegel Blog, 8 November, available at /blog-post/beyond-training-gap-learning-foundational-skills-job

OECD (2019) Individual Learning Accounts: Panacea or Pandora's Box? Organisation for Economic Co-operation and Development, available at 

OECD (2021a) Education at a Glance 2021: OECD Indicators, Organisation for Economic Co-operation and Development, available at 

OECD (2021b) ‘France’ in Education at a Glance 2021: OECD Indicators, Organisation for Economic Co-operation and Development, available at