Neither stack is better in the abstract. Each internalises different costs and externalises different risks, so the decision should not start with W3C-VC or SD-JWT. It should start with the credential. This page sets out a practical method drawn from the VC guide v1.1 (EUNIS SIG, July 2026), Verifiable Credentials in Higher Education: W3C-VC and SD-JWT VC, published by the EUNIS Special Interest Group – Mobility & Digital Credentials (EUNIS SIG): characterise each credential, see which profile it matches, and plan for two lanes rather than one winner.
One decision comes even earlier: what type of credential you are issuing (a badge, a formal microcredential, a degree) and under which governance. That question is covered in the decision framework for higher education, based on the Badges guide v1.1 (EUNIS SIG, July 2026), Badges and Digital Credentials in Higher Education. The type does not choose the format; the type’s requirements should.
Characterise the credential first: five dimensions
Ask these five questions of every credential type in your portfolio before any format conversation begins. Answered honestly, they make most of the decision.
| Dimension | The question to ask |
|---|---|
| Lifespan | Will this credential still be presented in ten, twenty or thirty years, or is it short-lived and frequently reissued? |
| Audience | Will it be checked by parties you have never coordinated with (foreign universities, employers abroad, unfamiliar regulators), or by members of a scheme whose governance you share? |
| Semantic density | Must its content keep the same meaning across systems (qualification levels, credit frameworks, learning outcomes, professional rights), or is it a few self-explanatory attributes? |
| Status complexity | Can it ever legitimately be “on hold” (reviews, corrections, proceedings), or is its lifecycle strictly issue-then-expire or issue-then-revoke? |
| Regulatory exposure | Is it inside the perimeter of eIDAS 2.0 and GDPR obligations that name specific technical properties, or outside it? |
When SD-JWT fits
The profile: short-lived, intra-scheme, semantically light, binary status, modest regulatory exposure. In a university, that covers:
- Student identity and enrolment status. Reissued every term, so cryptographic change is absorbed by natural turnover; verified mostly within coordinated national or campus ecosystems; semantically trivial.
- Access and affiliation proofs. Library, sports facilities, campus services, staff affiliation: short-lived, local, binary.
- Attendance and participation attestations where no formal recognition is claimed and the audience is known.
- Credentials inside a single governance scheme, where one authority defines the type and every verifier implements the same interpretation.
Here SD-JWT’s strengths are decisive: weeks to production, familiar JWT tooling, the deepest wallet support today (SD-JWT VC is one of the two formats the binding EUDI Wallet annexes name) and mobile-native OpenID4VCI/OpenID4VP flows. Deploying W3C-VC with full ELM alignment and SHACL validation for a library card would spend semantic infrastructure on an object that carries no semantics. That is poor engineering, and it is worth saying so plainly.
When W3C-VC fits
The profile: long-lived, cross-border, semantically dense, complex status, high regulatory exposure. Concretely:
- Degrees and diplomas. Thirty-year objects shown to strangers, carrying qualification frameworks, credits and learning outcomes that must mean the same in Tarragona and in Tallinn. The hypothetical worked example of Aina’s URV Master’s in Computing (EQF 7, issued 2026) is followed through the credential lifecycle.
- Formal microcredentials under the EU Council Recommendation, whose mandatory fields (EQF level, ECTS workload, quality assurance reference) are exactly the governed European semantics
@contextwas built to carry. - Professional qualifications linked to regulated practice, where interpretation drift becomes a legal dispute.
- Any credential that may need suspension, because a review process exists somewhere in its life.
- Any credential for which unlinkability is an explicit requirement, or whose privacy guarantees must be demonstrable as format properties rather than procedures (Article 25 GDPR, Regulation (EU) 2024/1183).
For these objects, the higher entry cost (DIDs, key infrastructure, ELM alignment, SHACL authoring, maturing wallet support for bbs-2023) is the price of properties the credential type actually requires.
Two lanes, not one winner
Put the two profiles side by side and the conclusion follows: most universities should expect to run both stacks. This is not a compromise; it is the correct reading of a portfolio that contains both library cards and diplomas.
| Fast lane | Formal lane | |
|---|---|---|
| Stack | SD-JWT | W3C-VC (VCDM 2.0) |
| Credentials | Student ID, enrolment, access, affiliation, attendance | Degrees, formal microcredentials, professional qualifications |
| Profile | High volume, short-lived, semantically light, coordinated verifiers | Long-lived, cross-border, semantically dense, regulated |
| Status | Token Status List (revocation only in the European profiles) | Bitstring Status List (revocation and suspension) |
The formal lane also has to reckon with the wallet rules. Since Implementing Regulation (EU) 2026/1731 (in force 11 August 2026), the binding annexes of the EUDI Wallet implementing acts name SD-JWT VC and ISO/IEC mdoc for person identification data and for attestations (ETSI TS 119 472-1 profiles). W3C VCDM is not among them today, but the regulation’s recitals state that wallets should also support it, as the reference format for attestations in education, once new W3C VCDM profiles are available. ARF v3.0 currently treats W3C VCDM as optional and for non-qualified attestations only. See EUDI Wallet.
The hybrid cell (VC-JOSE-COSE) is an enveloping option the W3C specifies: an institution can issue VCDM 2.0 credentials enveloped in SD-JWT and re-secure the same data model with Data Integrity and bbs-2023 later, keeping its semantic investment. Whether national wallets will accept it is not settled by the current acts, since a VCDM credential enveloped this way is not the IETF SD-JWT VC format the annexes name. The transition path also runs in one direction only: an SD-JWT credential cannot later be moved to another proof mechanism without reissuing it.
The classic mistake
The real risk is not “choosing the wrong stack”. It is picking a stack for the wrong reason, most often stretching the pilot’s stack to cover the diploma because the pilot was a success, and finding out ten years on that it does not fit. A pilot succeeds at issuance; a diploma has to survive its whole lifecycle, including the day its signing algorithm is deprecated. Framing the decision as winner-take-all guarantees that one class of credential ends up in a format that does not fit it, and it is usually the thirty-year class that loses.
A practical checklist
- Settle governance first. Legal basis, awarding authority, quality assurance reference and GDPR controller for each credential, before any format is chosen.
- Characterise every credential type on the five dimensions above.
- Get your Member State wallet roadmap in writing. Your students will hold what your national EUDI Wallet supports. Obtain its committed formats and proof suites as a versioned planning input, not as conference-corridor knowledge.
- Budget and staff for two lanes from the start.
- Price the lifetime and liability curve, not the issuance. Ask in writing what happens to every credential already issued when the signing algorithm is deprecated. If the answer is a reissuance campaign, put a number on it in the business case.
- Let regulation decide where it applies. Unlinkability, data protection by design, demonstrable accountability and suspension are obligations the institution already carries, not one criterion among ten in a scoring matrix that “ease of integration” can outvote.
EUNIS SIG position
This position applies to one case only: formal, regulated, long-lived European educational credentials. It does not apply to fast-lane credentials such as student IDs or access proofs, and it is not a claim that SD-JWT is deficient at what it was designed for.
For this class of credential, the regulatory framework names specific properties, and the SIG’s view is that only W3C-VC delivers them as properties of the format rather than as commitments of the controller. Four requirements, four correspondences:
- Unlinkability: Article 5a(16) of Regulation (EU) 2024/1183 prohibits post-issuance tracking and correlation and requires unlinkability where the user need not be identified; W3C-VC with bbs-2023 delivers it at the level of the proof (status indexes and credential identifiers must still be designed not to correlate), while SD-JWT approximates it through batch issuance, an operational workaround with documented limits.
- Data protection by design (Article 25 GDPR): BBS embeds data minimisation and SHACL shapes embed purpose limitation as technical properties; on the alternative stack both remain procedural.
- Demonstrable accountability (Article 5(2) GDPR): SHACL shapes registered in the EBSI Trusted Schemas Registry are immutable, dated, publicly inspectable artefacts, so compliance can be verified by inspection rather than by auditing application code.
- Suspension: European practice requires credentials that can be put on hold without being destroyed; Bitstring Status List provides that state, while the profiles deployed on the alternative stack do not.
In the SIG’s words, for these credentials choosing W3C-VC “is not a bet on a technology; it is the alignment of the credential’s format with obligations the institution already has”. Source: VC guide v1.1 (EUNIS SIG, July 2026).
Read against the wallet rules in force, the binding EUDI Wallet formats today are SD-JWT VC and mdoc, and Implementing Regulation (EU) 2026/1731 itself anticipates W3C VCDM support for education once new profiles exist. In the wallet, the formal lane therefore depends on those profiles arriving, and institutions should meanwhile keep an SD-JWT VC route for wallet delivery.
For institution-level advice beyond format choice, see the recommendations for higher education.
Key points
- Decide the credential type and its governance first, then characterise it: lifespan, audience, semantic density, status complexity, regulatory exposure.
- SD-JWT fits short-lived, intra-scheme, semantically light credentials; W3C-VC fits long-lived, cross-border, semantically dense, regulated ones.
- Most universities should run both stacks: a fast lane and a formal lane.
- The binding EUDI Wallet formats today are SD-JWT VC and mdoc; the regulation foresees W3C VCDM support for education once new profiles exist, so keep an SD-JWT VC route for wallet delivery meanwhile.
- The hybrid cell (VC-JOSE-COSE) is a one-way transition path the W3C specifies for the formal lane; whether national wallets will accept it is not settled by the current acts.
- Do not extend the pilot stack to the diploma just because the pilot went well; price the lifetime, not the issuance.
← Selective disclosure vs unlinkability · Credentials · Credential lifecycle →
