The costs of late quality signals in scholarly publishing
This study quantifies the significant scientific and financial consequences of delayed quality signals in scholarly publishing, revealing that journal discontinuations for quality issues often occur after substantial publication and financial commitments have accumulated, with publication outputs subsequently plummeting by nearly 80% within five years and exposing institutions to tens of millions of dollars in article-processing charges.
Original paper licensed under CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/). This is an AI-generated explanation of the paper below. It is not written or endorsed by the authors. For technical accuracy, refer to the original paper. Read full disclaimer
In the vast ecosystem of academic research, journals act as the gatekeepers and record-keepers of new knowledge. When a scientist publishes a paper, it is usually accepted into a specific journal, which then lists it in major databases that researchers around the world use to find information. These databases, such as Scopus, curate their collections carefully, deciding which journals meet high standards of quality and which do not. Occasionally, a journal that was once included is removed because its editorial standards have slipped or because it has become a vehicle for low-quality or predatory publishing. This removal is a formal signal to the scientific community that the journal is no longer trustworthy. However, there is a critical delay in this process. By the time a journal is officially flagged and removed from the list, it has often already published thousands of papers, and researchers have spent millions of dollars to publish them. The question remains: how much scientific work and financial investment accumulate in the shadow of a journal before the warning sign finally appears, and what happens to that work once the signal is finally given?
A recent study by Abdelghani Maddi, a researcher at Sorbonne Université in Paris, investigates exactly this timing gap. The researcher focused on 722 journals that Scopus had officially discontinued due to quality concerns. They linked these journals to a massive collection of over 1.1 million articles and reviews found in the OpenAlex database, a comprehensive index of scholarly work. By combining this publication history with records of actual payments made for open-access publishing, the researcher could measure the scale of the exposure. They found that the warning signal arrives far too late to prevent the bulk of the activity. Nearly one-quarter of the papers linked to these troubled journals were published after the year the journal was officially marked for discontinuation. This suggests that authors and institutions continued to invest in these journals for a significant period even after the decision to remove them had been made or was becoming known.
The study also tracked how the output of these journals changed after the discontinuation signal. The researcher compared the journals that were removed to a similar group of journals that remained in the database. In the year the discontinuation was recorded, the drop in the number of new papers was small and barely noticeable. However, the decline was not immediate; it was a slow, creeping process. One year after the signal, the journals were publishing about one-third fewer papers than their healthy counterparts. Two years later, the drop reached nearly 61 percent. By the fifth year, the output had plummeted by almost 80 percent. This pattern reveals that the scientific community does not react instantly to a quality warning. Instead, the adjustment happens gradually over several years, as authors slowly stop submitting to these journals and institutions stop paying for them.
Beyond the number of papers, the study quantified the financial cost of this delay. While the researcher could only see actual payment records for a small fraction of the articles, they used known price lists and statistical methods to estimate the total money spent. The most reliable evidence, based on direct price lists and observed payments, points to an exposure of tens of millions of US dollars. When they used broader estimates that included less certain data, the potential total rose significantly, but the researcher emphasized that the most solid numbers still place the cost in the tens of millions. This financial commitment was locked in before the formal signal could stop it. The study makes it clear that these figures represent the cost of the delay itself, not a judgment on the validity of every single article published in these journals.
The core finding of this work is a temporal asymmetry: the commitment of science and money builds up long before the formal quality signal arrives, and the correction takes years to unfold. The researcher cautions that a journal-level flag is a blunt instrument; it cannot tell us which specific articles are bad, only that the journal as a whole has lost its standing. Because the signal is retrospective, it cannot prevent the accumulation of work that happens in the meantime. The study suggests that the scientific community needs better, more transparent ways to track the status of journals in real time, rather than relying on these late, sweeping signals that leave a long tail of unresolved scientific and financial exposure behind.
Drowning in papers in your field?
Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.