Data Sharing Policy for Journals: FAIR Compliance Requirements
Reviewed by the NeucitePress Editorial Board — PhD academics, peer-reviewed journal editors and medical communication specialists.
Last updated: July 2026 • Reviewed by the NeucitePress Editorial Board • Reading time: 8 minutes
A growing share of academic journals now ask authors for a data availability statement, and a growing number require one as a condition of publication. This shift is driven partly by funder mandates, partly by the reproducibility concerns that have shaped research integrity discussions over the past decade, and partly by the FAIR principles, which have become the reference framework most journals and publishers use when writing their own data-sharing policies. This guide explains what FAIR means, why journals are adopting data-sharing requirements, the different policy tiers journals commonly use, where authors can actually put their data, and how an editorial team can draft a workable policy of its own.
What Are the FAIR Principles?
FAIR stands for Findable, Accessible, Interoperable, and Reusable. The principles were first articulated in a widely cited 2016 paper in the journal Scientific Data (“The FAIR Guiding Principles for scientific data management and stewardship,” Wilkinson et al.) and were designed less as a rigid checklist than as a set of goals for how research data should be managed so both humans and machines can find and use it. In brief:
- Findable: data (and its metadata) should have a persistent, globally unique identifier — typically a DOI — and be described with rich enough metadata that it can be located through standard search and indexing tools.
- Accessible: data should be retrievable using a standard, open protocol, with clear, documented conditions for access. Accessible does not mean every dataset must be free and unrestricted — it means the process for obtaining access, including any legitimate restrictions (patient privacy, commercial sensitivity), is transparent and follows a defined procedure.
- Interoperable: data should use formats, vocabularies, and structures that allow it to be combined and exchanged with other datasets and tools, rather than being locked in a proprietary or idiosyncratic format.
- Reusable: data should come with a clear usage license and enough contextual metadata (methods, provenance, variable definitions) that someone outside the original research team can understand and correctly reuse it.
A crucial and frequently misunderstood point: FAIR is not synonymous with “open.” A dataset can be FAIR while still being access-controlled, provided the conditions and process for access are clearly documented and consistently applied. This distinction matters enormously for sensitive data types, such as clinical or Indigenous data, where full open release is neither appropriate nor legally permissible.
Why Journals Are Adopting Data-Sharing Requirements
Several forces are pushing journals toward formal data-sharing policies rather than leaving it as an informal norm:
- Funder mandates: major funders increasingly require data management and sharing plans as a condition of grant funding, which creates pressure on journals to align their own requirements accordingly.
- Reproducibility and verification: reviewers and readers increasingly expect to be able to check reported results against underlying data, particularly for statistical or computational claims.
- Publisher-level policy: many large publishers have adopted baseline data-availability requirements across their journal portfolios, so individual journals inherit a policy rather than writing one from scratch.
- Reduced friction with structured statements: a standardized data availability statement, placed in the same location in every article, is far easier for the journal, reviewers, and readers to check than an ad hoc mention buried in the methods section.
Common Policy Tiers
Not every journal requires the same level of data sharing, and the differences matter both for authors deciding where to submit and for editors drafting a policy. Journal data policies generally fall into a few recognizable tiers:
| Tier | What it requires | Typical use case |
|---|---|---|
| Mandatory deposit | Underlying data must be deposited in an approved repository, generally at or before publication, with a persistent identifier cited in the article | Fields with strong reproducibility norms (genomics, structural biology, some social sciences); increasingly common at large open-access publishers |
| Data availability statement required, deposit encouraged | Every article must include a statement describing where and how the data can be accessed, but deposit in a specific repository is not always mandatory | The most common middle-ground policy across many disciplines today |
| “Available on request” | Authors state data will be shared upon reasonable request to the corresponding author, without a formal repository or persistent identifier | Common in fields with sensitive or proprietary data, though increasingly discouraged by editors and funders because it is difficult to verify or enforce |
| No formal requirement | No standard statement or policy; data-sharing left entirely to author discretion | Increasingly rare among established journals, more common in smaller or newer titles that have not yet formalized a policy |
“Available on request” statements have drawn particular scrutiny in the research-integrity literature because studies that have attempted to actually request such data have frequently found the data unavailable, which is part of why funders and journals increasingly push authors toward deposit-based policies instead.
Repository Options
Journals do not need to build or host a repository themselves; the standard approach is to point authors toward established, trusted repositories:
- General-purpose repositories: Dryad, Zenodo, Figshare, Harvard Dataverse, and the Open Science Framework (OSF) all accept data from any discipline, assign persistent identifiers (typically DOIs), and are widely recommended across publishers as a default option when no discipline-specific repository exists.
- Discipline-specific repositories: many fields have their own purpose-built repositories with domain-appropriate metadata standards — for example, GenBank and other International Nucleotide Sequence Database Collaboration archives for genetic sequence data, the Protein Data Bank for macromolecular structures, and ICPSR for social science survey data. Directories such as re3data.org help authors and editors identify the appropriate repository for a given data type.
- Institutional repositories: many universities operate their own repositories, which can be a reasonable option, though they sometimes lack the long-term preservation guarantees and discipline-specific metadata standards of dedicated data repositories.
A practical detail worth building into any policy: several data repositories, including Dryad, integrate directly with manuscript submission systems and support private, reviewer-only access to datasets during peer review, then make the data public alongside the published article — which resolves the common concern that early deposit means early public exposure of unpublished results.
How to Draft a Journal Data-Sharing Policy
Editorial teams writing or revising a data policy typically need to address the following elements:
- Decide the tier. Choose between mandatory deposit, a required data availability statement with encouraged (not mandatory) deposit, or another approach, based on the journal’s discipline norms and what its editorial board is prepared to actually enforce.
- Specify where the statement goes. Most journals place a “Data Availability” section immediately before the references, separate from the acknowledgments or funding statement, so it is easy for readers and indexers to find consistently.
- Name acceptable repositories, or point to a directory. Either list specific approved repositories or refer authors to a general-purpose directory like re3data.org so they can select an appropriate one for their data type.
- Address exceptions explicitly. Clinical data with patient privacy concerns, legally restricted data, and data involving Indigenous communities or other special governance considerations all need a documented exception path rather than a blanket requirement that does not fit the reality of the field.
- State the minimum acceptable content of the statement. A statement that only says “data available on request” without saying from whom or under what conditions is not usually considered adequate; require authors to name where the data are, under what license or access conditions, and how to obtain them.
- Coordinate with existing checklists. If the journal already uses discipline-specific reporting guidelines (such as those catalogued by the EQUATOR Network), align the data policy with what those guidelines already require rather than creating a conflicting parallel process.
- Communicate the policy clearly to authors and reviewers. Publish it on the “Author Guidelines” page, reference it in the submission checklist, and give reviewers explicit instructions to flag a missing or inadequate data availability statement during review.
None of this requires the FAIR principles to be applied perfectly on day one. Most journals adopt data-sharing requirements incrementally — starting with a mandatory statement, then tightening toward deposit requirements as authors, reviewers, and editorial infrastructure catch up.
