The Directory of Open Access Journals has started asking for something it used to take largely on trust: proof that a journal’s editors actually exist, that they agreed to serve, and that genuine peer review happens before the index label is granted. The shift follows a sharp rise in applications from journals built around paper mills, fabricated editorial boards, stolen academic identities, and renamed titles that reapply after removal. It changes what the phrase “indexed in DOAJ” means for authors, funders, research offices, and promotion committees.
DOAJ — the Directory of Open Access Journals, founded at Lund University in 2003 — has long been the closest thing open access publishing has to a whitelist. More than 20,000 journals from over 130 countries now appear in the index, alongside more than nine million article records. University open access policies, funder mandates, and tenure files frequently treat a DOAJ listing as a quality checkpoint. When that checkpoint no longer holds, the costs land on the people who check and the people who publish.
A research support librarian I’ll call Amara keeps the publication list for her university’s open access fund. A year ago she checked whether a journal appeared in DOAJ and moved on. Now she checks the date of the last re-evaluation, whether the editorial board page lists verifiable institutional affiliations, whether the contact address resolves to a real institution, and whether a journal has been removed and quietly re-emerged under a closely related title. Her spreadsheet has grown from one tab to nine. Her funded hours have not changed.
What the revised checks actually require
DOAJ’s published application guide now asks journals to show their workings, not just assert them. An applicant must provide a full editorial board with names and institutional affiliations, describe its peer review process in enough detail that the type of review is clear, state its open access licensing terms, publish a verifiable contact address, and demonstrate how articles will be preserved. DOAJ also asks for a journal’s ownership and publishing history when it suspects a title is a renamed version of a removed journal.
The index has always been selective, but the balance of proof has shifted. Before, a journal could describe a peer review policy and rely on an editor’s word. Now DOAJ’s volunteer editors and staff are more likely to contact listed board members, compare affiliations with institutional directories, and reject applications where editorial board pages are thin or copied. The criteria DOAJ publishes for applicants make that scrutiny explicit.
Some of the changes are quiet ones: a request for an ORCID iD is not mandatory everywhere, but boards that lack any traceable scholar identities now draw closer review. A journal that lists the same five people across forty titles in unrelated fields will be asked to explain the overlap. DOAJ has also made clear that listing in other indexes, including Scopus or Web of Science, does not guarantee its own acceptance.
The surge was not a mystery
Paper mills manufacture manuscripts and sell authorship on demand. The journals that publish those manuscripts need an appearance of legitimacy, and a DOAJ listing has become a valuable form of that appearance. When the Retraction Watch database passed 50,000 retractions, a large share of the underlying cases traced back to the same structure: a paper farm produced the work, a thin journal published it, and an index listing laundered the result. DOAJ tightened its checks because the same pattern reached its application queue.
Amara saw the pattern from the university side. A faculty member submitted to a journal that was in DOAJ, paid an article processing charge of just under 400 euros, and received a decision in eleven days. The journal disappeared from the index two weeks later. The published paper remained online, with no Erratum and no apology. The faculty member had followed the institution’s written policy. The policy had not required checking how recently the journal had been re-evaluated.
COPE, the Committee on Publication Ethics, has spent the past two years warning that paper mills adapt faster than editorial checklists. Its guidance on paper mill detection describes journals that look legitimate at first glance but fail the slow tests: contact addresses that cannot be answered, editorial boards that do not know they are listed, and review times too short to include real peer review. DOAJ’s new evidence requirements target the same failure points.
Why the old shortcut stopped working
Indexation has never been a guarantee of editorial quality. That point is easy to miss because universities often write policy as if it were. An earlier AcademicJobs article on journal indexation and quality assurance made the same distinction: an index is a directory, not an audit. DOAJ has always said so; what changed is the number of journals testing the gap.
The tightening of criteria is progress, and it deserves to be named as such. DOAJ has a public list of journals added and removed, an appeals process, and a machine-readable data feed that makes removals visible to systems that choose to read it. The problem is that very few university systems do read it. A tenure checklist written in 2019 still says “publish in a DOAJ-indexed journal” as a static phrase, not a continuous check. The criteria improved; the institutional software did not.
Amara’s nine tabs are a structure, not a preference. She built them because the university’s repository workflow does not flag removals. The fund will still reimburse an article processing charge if the journal was listed on the day of acceptance. It will not reimburse if the journal was removed before the invoice arrived, and the researcher would need to notice that themselves. The rule did not change anyone’s behaviour; it changed who carries the risk.
What the tighter criteria still leave unsolved
A global directory can raise its standards, but it cannot fund local verification. DOAJ relies on a modest paid team and a much larger pool of volunteer editors who review applications against the criteria. The index’s scrutiny is real, yet it ends at inclusion. Once a journal is listed, the next check depends on someone inside a university choosing to look.
There is also a timing problem built into the publishing workflow. A journal can be listed in DOAJ in February, removed in April, and continue to accept submissions under that old badge for months. Authors who rely on cached pages, old PDFs, or colleagues’ spreadsheets will miss the change. The only reliable check is the live DOAJ record, and the live record requires someone to look.
Early-career researchers carry the heaviest version of this burden. They are told to publish in open access journals, warned away from predators, and handed a short list of indexes that no longer serves as a static safeguard. The result is not fraud by authors. It is a mismatch between the precision of the index and the bluntness of the advice built on top of it.
What research offices can do now
The practical steps are small, but they fail without staff time. The Think Check Submit checklist remains the clearest starting point for authors, and it now matters twice: before submission and again after acceptance. For research offices, the more durable fix is to stop storing DOAJ status as a yes/no field and start storing the date of the query and the date of the last DOAJ re-evaluation.
- Check the live DOAJ record immediately before submission and again before paying an article processing charge.
- Subscribe to DOAJ’s public removal feed or set a weekly query for titles in your repository’s most recent deposits.
- Require reimbursement claims to include a screenshot of the live DOAJ record, not just the journal name.
- Fund a named staff member — even a few hours a week — to verify editorial board pages and contact addresses for journals used by each department.
- Add a field in the institutional repository that records whether the journal was indexed on the date the article was accepted.
None of these steps is expensive. The omission is more a budgeting habit than a technical failure. Institutions pay for repository software, open access funds, and research information systems. The small human verification layer that would connect those systems to the live index is usually treated as somebody else’s job.
The next number worth measuring
Amara now records one final column: whether a DOAJ-listed journal remained listed twelve months after acceptance. Most do. The exceptions are what the university should be tracking, because each exception represents a researcher who followed policy and still ended up in a disappeared journal. If institutions want to know whether their publishing advice is working, that twelve-month retention rate is the number to start with.
It would not require a new system. It would require one query against the repository’s published titles and the live DOAJ data. The data exist. The question is whether a university would rather know the answer or continue to treat an index entry as a finish line.
Photo by Alexander Grey on Unsplash