How Forecasts Become Education Decisions

Skills Assessment and Anticipation, or SAA, is often described as a forecasting exercise, but its real purpose is broader: to connect changing work with decisions about education, training, guidance, and workforce planning. This chapter examines how an SAA system chooses what to measure, combines incomplete signals, and moves from evidence to higher-education action. The practical test is whether a forecast is fit for the decision being made—not whether it looks precise.

That distinction matters because an occupation forecast is not automatically a skills forecast, and a skills forecast is not automatically a measure of what education supplies. Between a labour-market signal and a graduate’s capability lie classifications, standards, curricula, funding rules, institutional choices, workplace learning, and time. Alignment happens across those interfaces.

Start with the decision, not the data

An SAA exercise should begin by asking what decision it is meant to inform. A government deciding how many publicly supported study places to offer may need information by qualification or occupation. A provider revising learning outcomes needs more granular information about the skills and tasks within an occupation. A career service needs accessible information about routes into work, prospects, and possible skill development. These uses overlap, but they are not identical, so one measurement unit cannot serve them all equally well.

The OECD review of 17 national, whole-economy exercises found that 13 focused on occupations, 8 included skills directly, and 5 analysed qualifications. Occupations are common because they appear in many labour-market datasets, are easier to forecast, and can serve a broad audience. They can also be translated later into qualifications or skills. That convenience creates a boundary: changes in skill requirements inside a stable occupation may be missed when the occupation is treated as the object of change.

The translation is useful when it is visible. Ireland, for example, develops occupation-level results from its labour-market research unit into future skill needs through an expert group. Slovenia matches occupations to skills through ESCO, while Italy links occupational information to qualifications and an Italian skills taxonomy. These steps make a forecast more actionable, but a classification link is an analytic bridge, not direct evidence that every worker in an occupation has the listed competence or that every programme linked to it teaches that skill equally well.

A qualification-focused system answers a different question. Graduate tracking can show how people from particular programmes move into employment, what they earn, or whether they report that their skills are used. That information can help universities consider programme content and can help prospective students compare choices. It still does not reveal every skill used in a job, and a weak outcome may reflect regional demand, work conditions, selection into the programme, or the timing of the measurement. Qualification data are valuable precisely when they are not asked to answer a vacancy-level question by themselves.

The evidence sources form complementary windows. Administrative employment and labour-force data provide broad coverage of actual jobs and occupations, but they generally do not observe unfilled positions or describe skill content finely. Online job vacancies are more granular and frequent, and can reveal new skill terms or emerging occupations, but they show advertised demand and may overrepresent jobs that are routinely posted online.