Anticipating Skill Needs and Adapting Higher Education

1. Designing skills assessment and anticipation exercises

Abstract

Understanding the purpose of the Skills Assessment and Anticipation (SAA) exercise is the first step in designing an exercise that delivers useful results. The purpose of the exercise impacts the choice of the unit of analysis, the scope of the exercise and the time horizon. This chapter describes these basic features of SAA exercises, including possible options for each feature and relevant international examples.

SAA systems serve multiple purposes. They can be used to plan education and training provision at the Vocational Education and Training (VET), higher education and adult learning levels, to inform selective migration policies, to update the content of education, training programmes and occupational standards, and to support youth and adults in making education or labour market decisions, among other (see Chapter 4 for more details).

Each of these objectives has important implications for the design of SAA. For example, while exercises intended to support decisions on the number of study places may provide information by qualification or occupation, exercises designed to update education or training curricula and occupational standards need more granular information on skill needs. Thus, the first step to develop a SAA exercise is to determine the main purpose that it will serve.

Once the purpose of the exercise has been decided, the design process can start. This process should ensure that the features of the exercise allow to obtain the most useful information possible for its intended purpose. This chapter describes the basic features of SAA exercises – notably, the unit of analysis of SAA exercises, the coverage and the time horizon – including possible options for each feature and relevant international examples.

1.1. The unit of analysis

Table 1.1 highlights the units of analysis used in the SAA exercises reviewed for this report. It shows that most SAAs focus on occupations (13 out of 17), with about half of the exercises focusing on skills (8 out of 17), and a minority analysing qualifications (5 out of 17).1

Table 1.1. Unit of analysis by SAA exercise analysed

SAA exercise	Country	Occupations	Qualifications	Skills
Jobs and Skills Australia	Australia	✓
JobBarometer	Austria	✓		✓
Future Skills Program	Canada	✓		✓
OSKA	Estonia	✓
Skills Anticipation Forum	Finland	✓	✓	✓
Occupations 2030	France	✓
BIBB-IAB Qualification and Occupational Fields Projections	Germany	✓
IAB Skills Compass	Germany			✓
Graduate Career Tracking System	Hungary		✓
SOLAS Skills and Labour Market Research Unit	Ireland	✓
Excelsior Information System	Italy	✓
Skills and Labour Platform	Italy	✓	✓	✓
Education and Labour Market project (POA)	The Netherlands	✓	✓
Skills Demand for the Future Economy	Singapore			✓
Labour Market Platform	Slovenia	✓		✓
SEPE Occupations Observatory	Spain	✓		✓
Trends and Forecasts study	Sweden		✓

Source: Authors’ elaboration.

The overwhelming focus on occupations can be explained by: (1) results using occupations can be used for a larger range of policies and by a wider audience than results using other units of analysis, and, if needed, can be translated into qualification or skill needs at a later stage; and (2) the relatively simpler forecast of occupation needs relative to forecasting qualification or skill needs, given that occupations are readily available in multiple datasets.

Systems in which results are used for multiple purposes tend to focus on occupations, such as OSKA (Estonia) or Jobs and Skills Australia. Results from OSKA are used to determine education and training provision, but also to provide career guidance or to inform migration policies. Similarly, results from Jobs and Skills Australia have an impact on occupational standards, as well as on migration policy, among other areas (OECD, 2023[1]).

In addition, labour market needs are often forecast at the occupation level in a first instance, particularly when using quantitative methods and relying on labour market data. Information on needs by qualification or by skill can be added at a later stage by matching occupations to the corresponding qualification or skill requirements. This is, for example, the case in Ireland, where results at the occupation level of the SOLAS Skills and Labour Market Research Unit are later developed by the Expert Group on Future Skill Needs to determine future skill needs. Other examples include the Slovenian Labour Market Platform, which obtains information at the skill level by matching occupations to skills using ESCO, or the Italian Skills and Labour Platform, which matches needs at the occupation level to needs at the qualification level using administrative information that establishes a link between qualifications and occupations, and to skills using the Italian skills taxonomy, the Occupational Sample Survey (Indagine Campionaria sulle Professioni) from INAPP.

A small number of exercises look directly at labour market needs using qualifications. One example is the Hungarian Graduate Career Tracking System. This exercise obtains information on labour market needs by tracking the labour market outcomes of recent graduates using graduate tracker surveys and administrative data. The level of unemployment and wages for graduates of different qualifications compared to the average level gives information on the extent to which a qualification is needed in the labour market. Exercises relying on qualifications tend to focus on higher education qualifications only and aim at informing higher education provision decisions and prospective higher education students’ choices.

Finally, most exercises looking directly at skills rely on online job vacancy (OJV) information and estimate skill needs based on how often a given skill is mentioned in job vacancies, across the entire economy. This is the case for the IAB Skills Compass (Germany) or Skills Demand for the Future Economy (Singapore). However, in some cases, skill-based SAAs collect qualitative information, often at the sectoral or regional level. This is the case for the SEPE Occupations Observatory (Spain). Qualitative exercises focusing on skills can also be used to identify new skills, as in the case of the CARIF OREF example, described in Box 1.1.

Box 1.1. Determining new skills through Skills Assessment and Anticipation - CARIF OREF (France)

SAA exercises can be used to determine whether new skills are or will be required within an occupation or to identify new occupations. An example of a SAA exercise that identifies new skill requirements within occupations is the exercise implemented by the Network Association CARIF OREF (Training Resources and Animation Centre/ Employment-Training Regional Observatory) in France. This exercise was implemented in 2022-2023 and used qualitative methods to identify changes in skills and training needs due to the green transition across three regions for a set of occupations.

Source: OECD (2023[1]), Assessing and Anticipating Skills for the Green Transition: Unlocking Talent for a Sustainable Future, Getting Skills Right, https://doi.org/10.1787/28fa0bb5-en.

1.2. The geographical and sectoral coverage of the exercise

The geographical and sectoral coverage of the SAA exercise is also an important parameter to set (OECD, 2023[1]). For this report, only SAAs covering the full economy, including all sectors, and the whole country were selected (see Table 1.2). However, some exercises offer additional information by sector (9 out of 17), such as OSKA (Estonia) or the SOLAS Skills and Labour Market Research Unit (Ireland). To do this, some exercises, such as the SOLAS Skills and Labour Market Research Unit (Ireland) and POA (the Netherlands) start from aggregated results for the full economy and then develop sectoral results while others, such as OSKA (Estonia) work bottom-up, starting from a sectoral analysis which is later aggregated to obtain results for the whole economy.

Regarding geographical scope, by design, all exercises analysed in the context of this report cover the whole country. In addition, most exercises offer information at the regional level (12 out of 17), as shown in Table 1.2. Generally, exercises first produce forecasts at the national level, with results by region being developed at a later stage. However, this is not always the case. For example, in Spain, results are first developed at the Autonomous Community level and then aggregated to the national level.

Finally, SAA exercises can also have different scopes regarding the education or training level the SAA exercise provides information to. Generally, most SAA exercises provide information to all education and training levels that prepare individuals to work in occupations covered by the exercise, excluding primary and general secondary education. This allows to minimise costs, since a single exercise provides information to all relevant education and training systems. However, there are some exceptions. For example, the Skills and Labour Platform (Italy) provides information on qualifications provided by Higher Education Institutions only, while SANQ in Portugal, the Qualification Needs Anticipation System, provides information on VET qualifications.

Table 1.2. The coverage of the exercise by SAA exercise analysed

SAA exercise	Country	All sectors	Sectoral	National	Regional	All education and training levels	Particular education/ training levels
Jobs and Skills Australia	Australia	✓	✓	✓		✓
JobBarometer	Austria	✓		✓	✓	✓
Future Skills Program	Canada	✓	✓	✓	✓	✓
OSKA	Estonia	✓	✓	✓	✓	✓
Skills Anticipation Forum	Finland	✓	✓	✓		✓
Occupations 2030	France	✓		✓	✓	✓
BIBB-IAB Qualification and Occupational Fields Projections	Germany	✓		✓	✓	✓
IAB Skills Compass	Germany	✓		✓	✓	✓
Graduate Career Tracking System	Hungary	✓		✓		✓	Higher Education
SOLAS Skills and Labour Market Research Unit	Ireland	✓	✓	✓		✓
Excelsior Information System	Italy	✓	✓	✓	✓	✓
Skills and Labour Platform	Italy	✓		✓	✓		Higher Education
Education and Labour Market project (POA)	The Netherlands	✓	✓	✓	✓	✓
Skills Demand for the Future Economy	Singapore	✓	✓	✓		✓
Labour Market Platform	Slovenia	✓		✓	✓	✓
SEPE Occupations Observatory	Spain	✓	✓	✓	✓	✓
Trends and Forecasts study	Sweden	✓		✓	✓	✓

Source: Authors’ elaboration.

1.3. Time horizon

Another critical variable in SAA exercises is the time horizon of the analysis. Depending on the objectives of the exercise, the timeframe ranges from an assessment of current labour market needs to long-term projections (6+ years). In some cases, a combination of different time horizons is used to balance precision and strategic foresight.

As shown in Table 1.3, SAA exercises analysed for this report mostly forecast long term skill needs (8 out of 17), followed by current skill needs (6 out of 17), medium-term needs (4 out of 17) and short-term needs (1 of 17).

Table 1.3. The time horizon of the exercise by SAA exercise analysed

SAA exercise	Country	Current	Short-term (6 months-2 years)	Mid-term (2-5 years)	Long-term (6 or more years)
Jobs and Skills Australia	Australia	current			up to 10 years
JobBarometer	Austria			3 years
Future Skills Program	Canada				5-10 years
OSKA	Estonia				10 years
Skills Anticipation Forum	Finland				up to 2030
Occupations 2030	France				up to 2030
BIBB-IAB Qualification and Occupational Fields Projections	Germany				up to 20 years
IAB Skills Compass	Germany	current
Graduate Career Tracking System	Hungary	current
SOLAS Skills and Labour Market Research Unit	Ireland	current
Excelsior Information System	Italy	3 months		5 years
Skills and Labour Platform	Italy	3 months
Education and Labour Market project (POA)	The Netherlands				6 years
Skills Demand for the Future Economy	Singapore			2 years
Labour Market Platform	Slovenia		up to 1 year	3-5 years	up to 15 years
SEPE Occupations Observatory	Spain	current
Trends and Forecasts study	Sweden				15-20 years

Note: The horizon for projects implemented under the Future Skills Program (Canada) varies depending on the specific project.

Source: Authors’ elaboration.

Short-term SAA exercises tend to produce more accurate and reliable results, as they rely on labour market trends that are less likely to significantly change within a short period (OECD, 2023[1]). These exercises are particularly valuable for immediate policy interventions, such as designing training programmes for currently unemployed individuals or adjusting short-term workforce development initiatives to meet emerging labour market needs.

In contrast, long-term SAA exercises support longer-term strategic planning. However, long-term forecasts may be less accurate, as they are subject to unforeseen changes, such as natural disasters or wars. Some exercises use scenarios, such as Occupations 2030 in France, to forecast skill needs in multiple possible outlooks (OECD, 2023[1]). For these exercises to remain relevant, inputs should be updated, but exercises with long-term horizons tend to be updated less often than shorter-term ones (OECD, 2016[2]).

A few countries rely on multiple SAA exercises with different horizons, such as Estonia, which uses the Labour Market Barometer to project skill needs at a 6-month horizon and OSKA at a 10-year horizon, or Slovenia, where the Labour Market Platform provides information on skill needs in the short-, medium- and long-run. Combining both approaches enables policy makers to address immediate labour market needs while also preparing for future workforce transitions.

The frequency with which the SAA exercise is conducted, balancing the need for timely, up-to-date insights with the practical constraints of data collection and analysis, must also be considered.

Updating SAA results frequently, such as conducting the exercise annually, allows for continuous monitoring of labour market trends and ensures that the results remain as accurate and relevant as possible. Most exercises reviewed in the context of this report are updated annually, particularly in cases where quantitative methods are used, such as in JobBarometer (Austria) which uses OJV data, or the SOLAS Skills and Labour Market Research Unit. Excelsior (Italy) is the exercise with the highest frequency of those analysed, relying on a rotating panel of employers to offer short term skill need projections at monthly intervals.

Frequent updates can be particularly beneficial in rapidly changing labour markets. However, this approach comes with significant challenges, as frequent data collection and analysis can be resource-intensive, both financially and administratively. Additionally, labour market actors, such as policy makers, education providers, and employers, may struggle to react effectively to frequent updates, limiting the practical impact of real-time insights.

Conversely, a low-frequency approach, where the exercise is conducted every few years, reduces the cost and effort associated with data collection and analysis. It may also allow policy makers more time to process and respond to findings before new data is introduced. An example of a lower frequency approach is the Trends and Forecasts study (Sweden), which is implemented every 3 years.2 The downside of a lower-frequency approach is that sudden structural changes – such as technological disruptions – may not be captured in a timely manner, leading to outdated or less responsive policy recommendations.

Ultimately, the choice of frequency depends on the objectives of the SAA system, available resources, and the level of labour market volatility. Some exercises, such as OSKA (Estonia), adopt a hybrid approach, where comprehensive SAA are conducted at longer intervals, complemented by more frequent, targeted updates on specific sectors or emerging trends. In OSKA, skill needs are forecast for a few sectors annually, completing an analysis of skill needs for the full economy every 5 years. Box 1.2 below describes the OSKA Programme in more detail.

Box 1.2. The OSKA programme in Estonia

The OSKA programme forecasts labour market needs over the next ten years in Estonia. OSKA is implemented by the Estonian Qualifications Authority (Kutsekoda), and the implementation is overseen by a coordinating council consisting of 11 members from ministries, social partners, the Estonian Central Bank and the Estonian PES. OSKA’s methodology and results are also discussed with a panel of advisers, consisting of 40 members from ministries, social partners, think tanks and research institutions.

The OSKA skills forecasting system consists of three different exercises: Sectoral studies, general forecast reports and ad hoc thematic studies. All three exercises build on a labour demand and supply model at the national level, which uses administrative data to produce forecasts at the educational and occupational level.

• Sectoral studies forecast labour market needs over the next 10 years separately for each sector. The studies build on results from the labour demand and supply model and qualitative data from expert panels, which consist of 20-30 experts. Employers make up approximately half of each panel, while both educational institutions and policymakers make up a quarter each. The studies develop proposals for improving the sector and its institutions, which are reviewed and followed up after two years. Each sector is analysed every five to seven years.

• General forecast reports present labour needs forecasts for the whole economy at the national level. The general forecast reports combine insights from the labour demand and supply model with insights from the sectoral studies. The report is produced annually but every three years a more detailed report is produced. The reports also include specific thematic analysis, providing input on specific topics for education and training.

• Ad hoc studies provide information for specific policy needs. The ad hoc studies are commissioned by ministries or other stakeholders and approved by the OSKA Coordinating Council.

OSKA produces different resources to make results from their exercises easily accessible. For example, the Estonian Qualifications Authority produces short videos summarising the reports and condenses the main findings into two-page briefs. OSKA results are used in the Skills Compass, which is a career guidance tool developed by the Estonian Qualification Authority. Policy makers use OSKA results to determine the number of places in VET programs, and, for Higher Education programs, financing contracts between universities and the Ministry of Education require HEI to use OSKA results when establishing or modifying study programmes. OSKA results are also used for labour market and migration policy.

Notes

← 1.  For more information on the SAA exercises analysed for this report, please see Annex A.

← 2.  See https://www.scb.se/hitta-statistik/statistik-efter-amne/utbildning-samt-forskning-inom-hogskolan/analyser-och-prognoser-om-utbildning-och-arbetsmarknad/trender-och-prognoser-om-utbildning-och-arbetsmarknad/ and Statistics Sweden (2024[19]) for more information.