The Transferability Axis

A teaching model for transferability: how broadly a skill retains its value across roles and contexts

Technical and generic are useful categories, but they do not form a clean binary. MIRA therefore treats transferability as a continuous property of a skill: how broadly it can retain its value across roles and contexts.

This is document 2 of 2. Document 1, “The MIRA framework”, describes the framework itself: the main domains, how a single competence is described generically, the translation chain from job description to level, and the six levels. Together the two documents keep three questions apart. WHAT: what is the skill? Answered by MIRA’s competence structure and descriptions, in document 1. BREADTH: how transferable is it? Answered by the transferability axis, 0–100. DEPTH: how deeply is it held? Answered by MIRA level 1–6 and its mapping to the EQF. This document deals with breadth and depth.

1. What the two categories do not answer

Most competence models group skills as technical or generic. The categories are useful, and they are well established in practice. What they do not tell you is how broadly a skill can be used. A technical skill can be highly transferable, and a skill everyone calls generic can be tightly bound to one context. That is why the discussion so often ends in an argument about definitions rather than in shared understanding: the categories are being asked a question they were never built to answer.

The problem is amplified by the fact that the same groups have many names, and that those names carry different implied meanings:

When two people use different words for the same thing, or the same word for different things, the competence conversation loses precision. And the boundary between the groups is a decision rather than an observation — useful for sorting, but not a measure of anything.

2. The solution: one axis from 0 to 100

Alongside the categories we add one continuous axis, and it measures one thing only: breadth. At the far left sit skills whose value is bound to a single context — low transferability, strong context dependence. At the far right sit skills that retain substantial value across roles, disciplines and contexts — high transferability. Everything else lies somewhere in between. The axis is not a renamed technical/generic scale: technical and generic are typical associations with its two ends, not definitions of them.

The axis is divided into four zones. The zones are reading aids, not sharp categories — the transitions are gradual:

With this model we stop asking which box a skill belongs in, and start saying where it sits: “this skill retains value in few contexts”, or “this one carries almost everywhere”. The category still describes the skill; the axis states how far its value reaches.

FIGURE 1

The transferability axis

A continuous property of a skill: how broadly it retains value across roles and contexts

LOW TRANSFERABILITY

value tied to one context, tool or system

HIGH TRANSFERABILITY

value retained across roles, fields and contexts

0 25 50 75 100

TOOL-SPECIFIC 0-25

DOMAIN-SPECIFIC 25-50

BROADLY APPLICABLE 50-75

UNIVERSAL 75-100

THE TEST QUESTION

“Across how many other roles and contexts would this skill retain substantial value without being relearned?”

Follow-up: how much contextual adaptation would be required? Short rule of thumb: how many roles could use it without retraining?

The zones are reading aids, not borders. Technical and generic remain useful categories — they are not the endpoints of this axis.

Figure 1: The transferability axis. The zones are reading aids, not borders — the axis measures breadth only.

3. The test question

The position is determined by one question:

“Across how many other roles and contexts would this skill retain substantial value without having to be relearned?”

A second question sharpens the answer: how much contextual adaptation would be required? Little or none pushes the skill right; substantial re-learning pulls it left. Almost no other contexts: far to the left. Almost every context: far to the right. In a meeting the short form is enough — “how far does this skill carry?” — and two people assessing the same skill usually land close to each other.

Three supporting questions for when the answer is unclear:

• Does the skill survive the tool or system being replaced?
• Is it useful in an industry other than your own?
• Would you have to learn it again if you changed field?

4. Granularity: the level of description decides whether the score means anything

A transferability score is only meaningful when the skill is described at an appropriate level of granularity. “Spreadsheets”, “AI”, “communication” and “project management” are labels, not skills: each covers practices whose value reaches very different distances.

Basic spreadsheet use retains value in almost any office role and sits around 70. Advanced spreadsheet modelling is tied to particular analytical contexts and sits around 45. Both are “spreadsheets”. Score the label and you get a number that describes neither.

The working rule: describe the skill precisely enough that two people would recognise the same practice, then place it. If a score feels impossible to agree on, the usual cause is not disagreement about transferability but a description that is still too broad.

Transferability can only be assessed meaningfully when the skill is described at an appropriate level of granularity.

5. Examples

Example 1: Using the internet

In 1995, finding information online was a specialist skill. It required training, dedicated courses and often help from an IT department. On the axis the skill sat around 15 — tool-specific.

Then it started moving — not because more people learned it, but because it began to retain its value in almost any role. Whatever the job, the ability to find, judge and use information online carries over. Today it sits around 95: no job advert mentions it, because everyone assumes it. Spread was the visible symptom; the movement itself was in breadth of value.

The point: nobody redefined the skill from technical to generic. Its breadth of value changed, and the position followed. Wider adoption alone would not have moved it — a skill can be common and still context-bound.

Example 2: Spreadsheets

Spreadsheets followed the same path, but a shorter one. In the 1980s they were a specialist tool for finance and analysis, around 20 on the axis. Today basic spreadsheet use retains value in most office roles, around 70. At the same time the advanced end — data modelling, macros, complex formulas — has stayed behind on the left, because that value is tied to particular analytical contexts. One skill label therefore covers several positions, which is why granularity matters as much as the score itself.

Example 3: Using AI tools

Using AI tools currently sits around 35–40: it is in motion, practice is uneven, and the contexts in which it holds value are still expanding. Using the internet curve as a reference, this is roughly where internet use was in 1998. Note what would justify moving it right: not the number of people using it, but evidence that the same practice retains value across more roles without being relearned. The model therefore gives not only a description, but a basis for planning.

Examples that stay put

FIGURE 2

Skills move over time

A position is not fixed — it moves right when the skill retains value in more roles and contexts

Nobody redefined these skills. They kept their value in more roles — and the shift is measurable.

Spread is not the same as transferability. What moved these skills was retained value across roles, not the number of people using them.

Figure 2: Positions are not fixed. These skills moved right because they retained value in more roles, not because more people used them.

6. Why this matters

The position on the axis says nothing on its own about how long a skill holds its value, but it does point to what threatens that value. The two ends are exposed to different kinds of change.

This is a tendency, not a rule. Tool- and system-specific skills are more exposed to technological change: tools are swapped, systems are replaced, and the competence loses value even though the person has not changed. Broadly transferable skills tend to be more robust across job changes and shifts of context. But some narrow skills are remarkably durable — certified operation of specialised machinery sits at around 10 and has held its value for decades — and some broad ones change substantially. Transferability tells you where the value reaches, not how long it lasts.

That has three concrete consequences for an organisation:

1. Recruitment: the axis tells you how portable a requirement is, not how demanding it is to acquire. How hard a skill is to build is a question of depth — MIRA level — and is treated separately in part 2. Requirements far to the left are tied to your context and should be weighed against how quickly they can be established in post.
2. Training: skills on the left are more often affected when tools, systems and regulation change, while skills on the right can be reused more broadly over time. How much

maintenance and investment is required does not, however, follow from the position alone and must be assessed case by case.

1. Career paths: internal mobility is in practice about how much of a person’s competence sits to the right. The axis makes this visible instead of assumed.

The most important thing, however, is the language. When a leadership team stops debating which category something belongs to, and instead discusses how far its value reaches and which direction it is moving, the conversation changes character: from definition to decision.

7. How MIRA uses the axis

MIRA structures competence in four dimensions, and each competence is assessed across six levels that are mapped to the EQF levels. The transferability axis is a supplement to this — not a replacement, and not a substitute for the technical/generic distinction. It answers a different question from either.

The axis also solves a practical problem: where technical skills belong in the framework. The answer is that they are not a separate appendix. Technical skills are described with the same competence structure, the same level structure and the same assessment logic as every other skill — and, like every other skill, they are given a transferability score of their own. Some of them score high. That is the point: the framework keeps three questions apart. WHAT the skill is, BREADTH — how far its value reaches, and DEPTH — how deeply it is held.

Because MIRA is built on semantic analysis of competence descriptions, the position on the axis can be derived from how broadly a skill actually appears across roles and industries — and reassessed when the pattern changes.

The category tells you what kind of skill it is. Transferability tells you how far its value reaches — and that position moves.

From one axis to a multidimensional model

Breadth, depth and the three properties that change with level

8. Why one axis is not enough

The transferability axis answers how broadly a skill can be used. It says nothing about how well it is performed. Two people can hold the same skill, placed at the same point on the axis, and still be entirely incomparable in practice: one follows an instruction, the other sets the standard others follow.

The model therefore needs two axes, not one:

The decisive point is that the two axes are independent of each other. A narrow technical skill can be practised at level 6, and a universally applicable skill can be practised at level 1. It is precisely this independence that a binary classification cannot express — and it is the most common source of misunderstanding in competence conversations. It also means that how demanding a skill is to develop belongs to the depth axis, never to transferability.

9. The depth axis: the six levels

MIRA’s six levels describe the depth of practice. MIRA’s level descriptors draw on progression principles found in the EQF and related skills frameworks, particularly increasing autonomy, responsibility and complexity of practice. Each level is defined by these three properties read together: autonomy (how independent), responsibility (how far the responsibility reaches) and complexity (how compound the situations are).

The levels are described in full in document 1, “The MIRA framework”. They are repeated here in short form, because it is the depth dimension we need in order to read the grid later in this document:

10. The three parallel movements

What is interesting is that the levels do not simply mean “progressively better”. Three parallel movements take place, and they can be read separately:

Autonomy

guided → within boundaries → independent → developing → direction-setting → pioneering

Responsibility

own task → own work → shared result → development and quality → strategic area → systemic development

Complexity

familiar → somewhat varied → compound → complex → strategically complex → novel and systemic

This separation has practical value: it makes it possible to explain why two people who “know the same thing” are nevertheless placed at different levels. One works independently, but in familiar situations. The other handles novel and compound situations, but within boundaries someone else has set. The level is a combination of all three — not a score.

The movements also give a language for development conversations. Instead of “you need to get better” you can say: “you are confident in familiar situations, the next step is handling the compound ones”, or “you deliver your own work well, the next step is responsibility for shared results”.

11. The two axes together

Put the two axes together and you get a grid: transferability horizontally from 0 to 100, level vertically from 1 to 6. Every skill a person holds can be placed as a point. The four corners explain why this is more precise than two categories:

The grid also makes a common fallacy visible: that transferable skills are “soft” or easy. A universal skill practised at level 6 is among the hardest things to build, and takes the longest.

Figure 3: Breadth and depth are two independent axes. Every combination of position and level is possible.

12. Examples placed in the grid

The same examples from part 1, now with a level added. The level describes a concrete practice, not the skill in itself — the same skill exists at several levels:

Note that the columns do not track each other. That is the whole point: high transferability does not mean a high level, and low transferability does not mean a low level.

FIGURE 4

Examples placed in the grid

Transferability (0–100) horizontally, MIRA level (1–6) vertically. The two axes are read together.

The points scatter in every direction. Position says nothing about level, and level says nothing about position.

Figure 4: The same examples placed in the grid. The points scatter in every direction — position and level are independent.

13. What this gives the organisation

The two axes provide different kinds of decision support. Transferability shows how broadly the competence can be applied, and thus its potential for mobility and reuse. Level shows how advanced the practice has to be. Training, maintenance and investment needs must be assessed in addition.

1. Recruitment: the level says how advanced a requirement must be practised, and the position how broadly the requirement applies beyond your own context. Whether a requirement must be in place from day one or can be built internally is a separate judgement — it does not follow from the position alone.
2. Competence development: the level says how advanced the practice must be, and the position how broadly the competence can be reused afterwards. How often it must be maintained depends on tools, regulation and changes in practice, and is assessed separately.
3. Job descriptions: a role is described precisely when every skill has both a position and a level. It then becomes clear what actually separates two roles.
4. Career paths: internal mobility is about how much of a person’s competence sits both high and to the right. The grid makes this visible on a single page.
5. Risk picture: an organisation whose entire depth sits on the left holds little competence that can be reused across roles. The grid makes the pattern visible; how vulnerable it is must be assessed from how stable the tools and regulation actually are.

14. The link to MIRA and the EQF

The depth axis is not something new placed on top of MIRA — it is MIRA’s six levels. The EQF formally describes its levels through knowledge, skills, responsibility and autonomy, with complexity carried in the progression between levels. MIRA’s level descriptors draw on the same progression principles, read as increasing autonomy, responsibility and complexity of practice, and that is why MIRA’s six levels can be mapped to the EQF’s eight. An organisation can choose which of them to use as its reference without the content changing.

The transferability axis is an addition, and it is MIRA-specific. It does not describe the person, but the skill: how broadly it applies, and how it moves over time. Together this gives three questions that must be kept apart:

Because MIRA is built on semantic analysis, both axes can be derived from the language of the competence descriptions: the breadth from where the skill appears across roles, the depth from how autonomy, responsibility and complexity are formulated.

For a full walkthrough of the framework's structure — the main domains, the generic competence descriptions and the translation chain from job description to level — see document 1, “The MIRA framework”. The two documents are made to be read together: the first explains what the competence is, this one explains how broadly it applies and how deeply it is practised.

15. Summary

Technical and generic describe useful perspectives on skills. Transferability answers a different question: how broadly a skill retains its value across roles and contexts. Level answers a third: how deeply it is held. You have to read all three.

Part 1 made breadth explicit and continuous, assessed independently of any category. Part 2 added depth, and showed that breadth and level are independent quantities. MIRA does not claim that the established categories are wrong; it makes explicit a property they do not capture

precisely enough. The result is a model with three layers — WHAT the skill is, BREADTH: how broadly it retains value, and DEPTH: how deeply it is practised — where each layer answers its own question, and none of them makes the others redundant.

16. MIRA in relation to SkillsFuture and the OECD

MIRA does not stand in opposition to established competence frameworks such as SkillsFuture Singapore or the OECD's work on skills. The model is better understood as a supplementary layer that makes certain properties of skills more explicit.

SkillsFuture Singapore distinguishes, among other things, between Technical Skills and Competencies (TSC), which are tied to particular occupations and job roles, and Critical Core Skills (CCS), which describe skills with broad application across working life. That division is particularly useful when skills are linked to job roles, training and career pathways. At the same time, technical skills too can be transferable between roles and sectors. “Technical” and “transferable” are therefore not opposites.

The OECD applies several perspectives on skills, including technical, transversal, information-processing, and social and emotional skills. Its work on transversal skills also addresses the degree of transversality — the extent to which a skill can be applied independently of a particular job or task. OECD analyses likewise show that skills with high transversality are not only social or so-called “soft skills”; technical and digital skills can also be highly transversal.

This is conceptually close to MIRA's understanding of transferability, but MIRA makes an operational move of its own:

MIRA treats transferability as an explicit continuous dimension from 0 to 100, and keeps that dimension separate from both skill type and skill level.

A technical skill can therefore have high transferability if it retains substantial value across many roles and contexts. Equally, a narrow, domain-specific skill can be practised at a very high level.

Three different questions

What MIRA adds

MIRA's most important contribution is that these questions are kept apart. A skill can be technical and broadly transferable at the same time. It can be narrow and still be practised at expert level. And the same overarching competence can be expressed differently in different working contexts.

This makes the model particularly relevant when the purpose is to analyse competence mobility, career transitions and which competence can be reused across roles and contexts. MIRA also has a practical strength in that transferability and level can be analysed semantically from concrete competence and job descriptions. That makes it possible to connect a stable competence language to a labour market in constant change.

The limitation of the model

Transferability is not a directly observable quantity with one objectively correct value. A score from 0 to 100 is a modelled judgement. The result is influenced by, among other things:

• how precisely the skill is described
• which roles and contexts it is compared against
• what evidence base the assessment draws on
• how the labour market and the use of the skill develop

The score should therefore be read as decision support, not as a universal or permanent measurement. That is also why granularity, and the ability to reassess a skill's position over time, are central parts of the model.

MIRA's position

MIRA does not replace SkillsFuture, the OECD's skill concepts or the EQF. These frameworks partly answer different questions and were developed for different purposes. SkillsFuture shows how technical and core skills can be structured operationally around job roles and skills development. The OECD's work provides a broader research and policy perspective on transferable and transversal skills, among others. The EQF provides an established reference for level and progression.

MIRA's contribution is to make the degree of transferability explicit and continuous, and to keep it separate from both what kind of skill it is and how advanced the practice is. MIRA can therefore be described as conceptually aligned with established work on transferable and transversal skills, while introducing its own operational model for representing transferability as a continuous dimension.