ORCID Research Assessment: Five National Models

ORCID research identification — linking a researcher’s persistent iD to the outputs they submit for evaluation — is no longer a REF-only story. Australia’s Excellence in Research for Australia (ERA), Italy’s Valutazione della Qualità della Ricerca (VQR), and United States federal disclosure rules under NSPM-33 all use the same underlying identifier to cut duplicate reporting and improve attribution accuracy.

ORCID (Open Researcher and Contributor ID) is a non-profit, community-governed registry that issues a free, persistent researcher unique identifier, used to disambiguate individuals and link them reliably to their scholarly outputs across institutions, funders and countries.

This is a systems-level comparison, not a REF compliance checklist. It sets out what “orcid research” actually means for national assessment infrastructure, and what research administrators in other jurisdictions can learn from five different implementation models.

• What is ORCID research identification, and why does it matter for assessment?
• How do national research assessment systems use ORCID?
• What measurable benefits has ORCID delivered so far?
• Frequently asked questions
• What should research administrators do next?

What is ORCID research identification, and why does it matter for assessment?

ORCID assigns each individual a 16-digit iD that stays constant across name changes, institutional moves and career stages. That persistence is what makes it useful for assessment exercises: a system built on ORCID iDs can match a researcher to their outputs automatically, instead of relying on manually typed names that are easily duplicated, misspelled or confused with a namesake.

For a searcher asking what is ORCID iD in research: it is the identifier layer that sits underneath a growing number of national reporting workflows, connecting a researcher’s ORCID record to journal articles, datasets, grants and peer reviews via APIs held by publishers, funders and institutional repositories.

Two problems drive adoption in assessment contexts:

• Reporting burden. Researchers and administrators re-key the same publication lists into multiple systems — institutional repository, funder portal, national assessment platform — for every reporting cycle.
• Attribution accuracy. Common surnames, transliteration variants and institutional affiliation changes make name-only matching unreliable at national scale.

How do national research assessment systems use ORCID?

Five jurisdictions illustrate distinct implementation models, ranging from “recommended” to a designated statutory disclosure identifier.

The common pattern is “enter once, reuse often”: a researcher curates one ORCID record, and every downstream system — grant portal, institutional repository, national assessment platform — draws from that single source rather than requesting a fresh manual submission.

What measurable benefits has ORCID delivered so far?

Attributed, publicly reported figures show the effect at scale in at least two of the five systems above, plus the underlying registry itself.

• The Australian Research Council reports that its 2018 ORCID integration into the Research Management System saw more than 1.4 million research outputs uploaded to researcher profiles, with roughly 940,800 of them imported automatically via ORCID across more than 14,000 researchers.
• ANVUR’s policy for the Italian VQR 2020–2024 requires participating researchers to register an ORCID iD and link it to submitted publications, explicitly to reduce duplicate reporting between institutional IRIS repositories and the national exercise.
• Under NSPM-33, US federal agencies including NSF and NIH require biosketch and current-and-pending-support disclosures through SciENcv, which requires a linked ORCID account — standardising researcher disclosure across agencies that previously used incompatible CV formats.
• The ORCID registry itself had issued more than 21 million iDs and counted over 1,400 member organisations — publishers, funders, universities and consortia — by 2024, giving national systems a large, interoperable base to build on.
• Research England’s REF 2029 open-access policy, which takes effect for outputs published from 1 January 2026, treats ORCID registration as good practice supporting output management, though it stops short of a mandatory requirement.

The comparison is instructive: jurisdictions that moved from “encouraged” (Australia, Finland, REF) to “required or designated” (Italy, US federal agencies) report the clearest reduction in duplicate manual entry, because auto-population only works reliably once linkage is near-universal across the researcher population being assessed.

Frequently asked questions

What does ORCID mean in research?

ORCID stands for Open Researcher and Contributor ID. In research, it is a persistent, free identifier that distinguishes one researcher from another with a similar or identical name, and links that person reliably to their publications, datasets, grants and peer reviews across institutions and countries.

Is ORCID free to use?

Yes. Individual ORCID registration and record use are free and always will be under ORCID’s governing principles. Institutional and publisher ORCID membership — which funds the non-profit registry and enables API-level integrations such as auto-population — is a paid tier, but it carries no cost for the individual researcher.

Is ORCID trustworthy?

ORCID operates as a non-profit registry governed by its member organisations, with published transparency and open-data principles. Researchers control what appears on their own record and who can see it, which is why national assessment bodies including ANVUR and the ARC treat it as a reliable base layer rather than a proprietary vendor system.

How to get ORCID research?

Register at orcid.org/register, a process that takes under a minute and requires only a name and email address. Once registered, a researcher connects the iD to institutional, funder and publisher systems so outputs and affiliations populate the record automatically for future assessment cycles.

What should research administrators do next?

The REF 2029 experience is one data point, not the template. Systems that made ORCID linkage a condition of participation — Italy’s VQR, US federal SciENcv disclosure — report faster convergence on clean, deduplicated researcher-output data than systems where linkage remains optional.

For institutions operating across multiple national or funder reporting regimes, three implications follow:

• Treat ORCID linkage as reporting infrastructure, not a one-off registration task — it must be maintained across staff transitions and repository migrations to keep auto-population accurate.
• Where a national exercise (or a funder mandate) has moved from “recommended” to “required,” expect the sharpest drop in manual re-keying, based on the Australian and Italian evidence above.
• Pair identifier infrastructure with contribution-level attribution standards: ORCID answers “who,” while frameworks such as the CRediT contributor role taxonomy answer “did what.” CASRAI originated CRediT in 2014; it is now stewarded by NISO as ANSI/NISO Z39.104-2022. Institutions building assessment pipelines benefit from aligning both layers rather than treating identifier and attribution separately — see the CASRAI overview of CRediT contributor roles and the wider research administration resources for related standards.

National research assessment is converging on a shared identifier layer even where the assessment models themselves differ sharply — peer review in Italy, metrics-assisted auto-population in Australia, statutory disclosure in the United States. The REF is one implementation among several, not the reference design.