Not a Good Year for Research Integrity

Last year a disheartening record was set for research integrity in scholarly publishing. In 2023, over 10,000 research articles were stamped “retracted” reducing them to ghost research. The previous year the number was short of 6,000, itself a retraction record. Clearly something is amiss in the quiet halls of academe.

When a paper is marked retracted, or rather RETRACTED, on page after page, the journal’s editors and publishers have determined that it has a problem well beyond “correction” or “update.” Corrections, for example, will be issued for a number of Claudine Gay’s articles, as the former Harvard president adds missing citations to work quoted (and as has been posted for a paper of mine, after I corrected the reversal of two books listed in a table).

Retraction, on the other hand, indicates flaws, if not deceptions, that undermine the integrity of science in ways (with details below) that have escaped the diligent attention of the editorial and peer review process. The retracted paper is not deleted, as its continuing presence oddly contributes to the integrity of the scholarly record (even as the paper’s publication undermines that integrity).

Among the more notorious cases is a 1998 Lancet article by Andrew Wakefield and 12 colleagues, reporting an association between childhood vaccinations and autism. In the years following publication, the study was discovered to have been highly compromised, with funding from lawyers whose clients were suing vaccine makers (who, in turn, had financial relationships with researchers attacking the Wakefield paper), and biased participant selection and invasive treatments. It took Lancet’s editors 12 years to retract the paper, following the revoking of Wakefield’s U.K. medical license for misconduct. In the meantime, he became a leading anti-vaxxer. While the paper was cited 1,370 times before retraction (and 2,960 times after), its 100,000 social media comments, not to mention Robert F. Kennedy Jr.’s presidential campaign, attest to the potential for a retracted paper’s worrisome afterlife.

Retraction marks the distinctive intellectual property (IP) in which scholarly publishing trades. These papers’ violations and dishonesty may fall outside the scope of IP law, but they are still very much a part of this industry’s global IP economy, which been valued at $20 billion annually (not far behind the recording industry but far more profitable, if for publishers alone, as the only ones being paid for this IP).

While the 2023 retractions included papers by the Canadian-born scientist Marc Tessier-Lavinge, who resigned from Stanford University’s presidency as a result in August, 8,000 of the retracted articles involved the open access publisher Hindawi. The company brings us to the economics of retraction. In 2020, Hindawi made $40 million (USD) through the article processing charges (APC) paid by authors’ grants and institutions for open access publication. It seemed particularly good at executing what is now the common commercial model for open access, and the mega-publisher Wiley acquired the company for $298 million in 2021, praising its “highly efficient publishing platform.”

Wiley soon discovered, however, that Hindawi journals, with their many special issues (boosting publishing volumes), were riddled with problem papers. Hundreds of “guest editors” were scamming the peer review system, helping authors publish already published papers, works produced by for-hire “paper mills,” and other illegitimate articles. Editors and authors were boosting their career opportunities, with some gaining bonuses for publication. Despite their’ blatant breaches of scholarly integrity, Wiley is having to explore a legal means to notify other publishers about the bad actors, as personal reputations appear to be afforded greater legal protection than scientific integrity.

And while Wiley, in response, has also expanded its “research integrity team” and announced an end to the Hindawi brand (a commercial retraction?), you can see how Hindawi’s increased publication rates drew Wiley to the company and how such rates will continue to act as a force in the IP economy of scholarly publishing. In April of last year, for example, Wiley fired the Journal of Political Philosophy’s long-standing editor and professor Robert Goodin for his unwillingness to “increase the number of articles we publish by a factor of 10,” according to Anna Stilz, one of the journal’s editorial board members. To their credit, more than a few of the academics who serve the corporate publishers as editors and board members have resigned (i.e., retracted their services) from leading journals out of a concern for the role played by the profit motive.

Still, the industry is taking steps to combat the harm and abuses leading to the actual retractions. There’s the STM Association Integrity Hub and AI-powered fake-paper detector. As for my work with the Public Knowledge Project, we’re developing a Publication Facts label (based on the nutrition facts model) for research articles, discussed in an earlier column. It may take another record year or two, unfortunately, but I do believe we can do a great deal to clean up this new publishing economy in the name of research integrity and public access. And tempering corporate impatience with scholarship’s pace is going to prove no less key to preserving that integrity.