The numbers don’t lie. Scientific publishing has a problem, and it’s getting worse. Vigilance against fraud and flawed research has always been necessary, but in recent years the sheer volume of questionable material has threatened to overwhelm publishers.
We were not the first to write about scientific misconduct and academic publishing issues when we launched Retraction Watch in 2010 and set out to cover the topic on a regular basis. At the time, coverage of these issues was sporadic, with gaps of a year or more between major scandals that merited national or international attention, such as the debacle of Italian surgeon Paolo Macchiarini ten years ago, whose misconduct (including the implantation of an artificial trachea) led to a European prison sentence and a documentary series on Netflix.
Today, reporters compete for scoops on scientific misconduct that regularly appear on the front pages of major news media around the world. In the course of the year, the presidents of Harvard and Stanford universities resigned amid allegations of research misconduct and plagiarism, and the Dana-Farber Cancer Institute demanded six retractions and even more corrections.
But it’s not just big-name universities that are being targeted: In the past few months, presidents of two other universities have resigned after we reported on allegations about problems with their jobs.
As for the rampant issue of paper retractions, over 10,000 paper retractions last year were the result of the activities of paper publishers, mainly dominated by one publishing house, Hindawi, a subsidiary of Wiley. Paper publishers are shadowy companies operating in countries such as China and Latvia that advertise their services on social media, selling researchers’ entire papers, authorship pages, and citations, and exaggerating their importance.
The total number of retracted papers is about 1 in 500 published papers, up from about 1 in 5,000 20 years ago, and while many of these papers concern difficult topics, the number is undoubtedly small compared with the amount of problematic research that exists.
So what’s going on? Are scams on the rise or are steps finally being taken to crack down on them?
Only a lengthy and costly audit of the literature over several decades can determine whether misconduct is truly becoming commonplace, but what is clear is that research misconduct has been industrialized on an unprecedented scale.
Paper mills in particular are seizing the opportunities presented by an academic culture of “publish or perish.” Publishers have been aware of these practices for at least a decade but have largely ignored them, but their thirst for growth and profits (Elsevier alone made about $2.2 billion last year) has led them to tolerate and even encourage the activities.
Mills seems to have hit a blind spot for big publishers: the number of papers published each year in lucrative but poorly monitored special issues has ballooned to nearly 3 million. Many of these papers, including some truly bizarre ideas like linking aerobics and dance training to geology, are never cited by other researchers and are probably never read by anyone. (And with the rise of generative AI, the proportion of papers that aren’t written by humans is probably increasing.)
Only recently, as mainstream journalists have become interested in paper mills and the questionable activities associated with them, have major companies like Elsevier, Springer Nature, and Wiley adopted the victim role and acknowledged the existence of paper mills, rather than acknowledging their complicity in creating the business models and incentives that fostered this behavior. Meanwhile, the paper mills bribed journal editors to publish their clients’ papers. These and other misconduct continue to plague the literature today.
Thankfully, a small army of volunteer detectives is tasked with bringing problematic papers to the world’s attention. But they risk being sued by those they investigate if they make their findings public. Still, public scrutiny seems to be the only way to bring the problem to the attention of publishers.
The public may believe that scientific fraud and sloppiness are the result of big pharmaceutical companies promoting fraudulent research in order to drive profits, a perception reinforced by cases like that of Cassava Sciences, whose experimental Alzheimer’s drug has been plagued by allegations of fraudulent data.
Meanwhile, some critics accuse scientists of bending the truth to get government grants — a theme in the fierce debate over the origins of the coronavirus that has caused the pandemic — and that producing results that satisfy funders and advocates can be key to success.
In practice, however, retractions of papers reporting such studies are far less common than retractions of purely academic studies. Drug-sponsored studies are heavily scrutinized by regulators, and the scrutiny of the data tends to discourage would-be falsifiers. (An entirely separate but equally important issue is the drug companies’ habit of hiding research that is unfavorable to the products they test.)
A more important reason researchers cut corners or cheat is bureaucracy: rankings. University rankings such as Times Higher Education rely heavily on the number of citations to papers published by an institution’s researchers. But citation accumulation is relatively easy to manipulate: the more papers a researcher publishes, the more they are cited. And under the current business model, where authors pay journals anywhere from a few hundred to more than $10,000 to publish their papers and make them available without subscription fees, publishers make more money. When a university improves its ranking, it attracts more top students and faculty, which in turn increases funding.
Finding and flagging fraud is essential work, but it’s like building more and more sewage treatment plants at river estuaries to stop ocean pollution: we urgently need to build more “treatment plants,” but we also need to shift the incentives (in this case, expose or disappear) so that waste doesn’t end up in rivers far upstream.
Here are three potentially far-reaching improvements: First, give more power and funding to government agencies like the Office of Research Integrity, which oversees research funded by the National Institutes of Health; Sen. Bill Cassidy (R-Louisiana) proposed this last month. Second, end the all-too-easily corrupt practice of counting the number of citations of a paper by other researchers as a measure of its quality. Finally, scientific journals need to end their paid-for business models that effectively sell a semblance of legitimacy by charging researchers publishing fees, no matter how painful this may be.
Science is being polluted and science must fight back.
