I have been working on this post for a while. I thought I'd drop my reflections. Academic peer review is still sold to us as the bedrock of research quality control. In principle, that is true. In practice, the system around it is beginning to look less like a collective scholarly endeavour and more like an extractive bargain. Researchers are asked to review more, submit more, format more, fix more, and tolerate more bureaucracy, all while publishers charge thousands in article processing fees, enjoy healthy margins, and steadily push more of their own work back onto the academic community.
I now receive around five review requests a week. That is not unusual. Many colleagues could say the same; and some get more. Reviewing has become a permanent background task layered on top of teaching, grant writing, administration, pastoral duties, supervision, and our own research. Yet the assumption remains that we will continue to absorb this work without complaint because peer review is “part of being an academic”. Perhaps it is. But that does not mean the current model is fair, efficient, or sustainable.
This post is not an argument against peer review. It is an argument against the way publishers and journals increasingly depend on unpaid academic labour while reducing their own responsibilities. I am getting close to boiling pontoon these topics; enough is enough.
The burden keeps rising, but the rewards do not
The number of submissions across many fields has grown rapidly, and the burden of reviewing has grown with it. The pool of suitable reviewers has not kept pace. The same people are repeatedly asked to assess manuscripts, often at short notice, often outside their immediate specialism, and often with a turnaround time that bears no resemblance to the reality of academic work (within 14 days!).
This creates a simple and predictable problem: reviewer fatigue. Many researchers now decline most invitations. Others accept only when the paper is especially close to their interests, or when guilt gets the better of them. In my own case, I often only accept reviews that genuinely interest me or when I have a rare block of dead time, such as before a long-haul flight. That is not because I do not value peer review. It is because there are limits to how much free labour any of us can continue to provide. And also my University does not value peer review, or the process.
And it is free labour. Reviewing takes time, attention, and expertise. A proper review is not a quick skim and a few throwaway comments. It involves reading carefully, checking methods, assessing claims, judging novelty, thinking through limitations, and offering constructive feedback. That is skilled intellectual work. Yet the standard reward is still an automated thank-you email and, if you are lucky, an invitation to review the next manuscript a week later.
This might be easier to defend if journals were obviously reinvesting in the parts of publishing that academics cannot do alone. But increasingly the opposite seems true.
Publishers are offloading more of their own work
One of the most frustrating shifts in academic publishing is that journals have become more demanding at the point of submission while appearing less willing to do the editorial and production work that publishers used to handle themselves.
We are now routinely asked to submit manuscripts in near-publication-ready form before the paper has even been seriously considered. That means carefully formatted references, tightly specified section structures, word counts, title page details, declarations, checklists, file naming conventions, figure placement rules, supplementary files, and cover letters tailored to each journal’s preferred style and tone. If the paper is desk rejected, much of that work has achieved precisely nothing. We then repeat the exercise for the next journal, and the next.
This is an extraordinary waste of academic time. It is also a quiet transfer of labour. What used to be part of the publisher’s production pipeline has increasingly become the author’s problem, and the author is doing it before acceptance rather than after.
I was reminded of this very recently while preparing a manuscript for PLOS Digital Health. I was asked to make extremely basic stylistic adjustments, such as changing “Figure” to “Fig”. That may sound trivial, and in one sense it is. But that is exactly the point. These are copyediting-level changes. They are not questions of scientific merit. They are not major reporting problems. They are not methodological flaws. They are house-style preferences.
Why, in 2026, are researchers being pushed to spend time on this at such an early stage, particularly when universities such as mine have already signed transformative agreements that cover open access publication? If institutions are already paying publishers through these arrangements, it is fair to ask what service is actually being bought. Because from the author side, it often feels as if the answer is: not very much.
This is part of a broader trend. Journals increasingly want polished, compliant, low-friction submissions, while reserving the right to reject them before review for reasons that are sometimes barely distinguishable from admin. It makes editorial workflows easier for the journal, but it does so by making submission more burdensome for researchers.
High fees, record profits, and unpaid goodwill
This would all be irritating enough if publishing were a fragile public good run on minimal margins. But that is not the reality. Much of academic publishing remains highly profitable. Researchers write the papers, reviewers assess them, editors are often academics too, and universities then pay to read the outputs or to make them open access. In other words, the academy produces the content, performs the quality control, and then pays again for dissemination.
Against that backdrop, the lack of meaningful reviewer reward is increasingly hard to justify. Article processing charges continue to rise. Transformative agreements spread the cost in a different way, but do not make it disappear. Publishers still get paid. The people doing the reviewing still do not. Take note of this point; those who generate the content; do the reviews; do not get paid. The journal does. Yet this blog is run on a value of £50 per year. Yet I am charged £5,000 for an article.
This is the central absurdity of the system. The most important quality assurance function in scholarly publishing depends on unpaid academic labour, yet the business model surrounding it is anything but charitable (they may suggest otherwise). Publishers frame peer review as a professional obligation when they need free expertise, but treat publication as a premium service when it is time to invoice a university or funder.
We are told there is no money to compensate reviewers. But somehow there is always money for platform development, branding exercises, publisher showcases, and complex submission systems that seem designed to maximise friction rather than reduce it.
JMIR shows both the promise and the problem
JMIR is an especially interesting example because it is one of the few publishers that has at least tried to reward reviewers in a tangible way.
Its old Karma Credits system was, for a time, one of the better ideas in this space. Reviewers earned credits that could be used against future publication costs. It was not perfect, but it recognised a basic principle that much of the publishing industry prefers to ignore: peer review has value, and reviewers should receive something more than polite gratitude.
That scheme then disappeared.
And this is where the story becomes revealing. Yes, JMIR is now offering author credit again (3 years after they ended it). But it did not for a good while, and there was never a satisfactory explanation of why it vanished in the first place (suggestion of the blockchain). The disappearance of a meaningful reviewer reward was treated almost as an administrative footnote. There was no serious public reflection on what had been lost, no real discussion of why the academic community should accept the withdrawal of something that had genuine value, and no convincing replacement during the period of absence.
There was also a lot of earlier rhetoric around blockchain and more ambitious reward models. At one point, there was talk of cryptocurrency-style extensions and a broader rethinking of how reviewer contributions might be recognised or exchanged. That conversation seems to have quietly evaporated. The blockchain idea drifted away with very little follow-up, and with no meaningful account of what happened to that vision.
Now that reviewer credit has returned, it feels difficult to greet it with uncomplicated enthusiasm. Of course it is better than nothing. Of course it is welcome. But the reintroduction is a little hollow.
Why? Because for many UK academics, the practical value of that credit is weaker than it first appears. If your university has already signed a transformative or publishing agreement that covers APCs, then credit against publishing charges is not necessarily much of a reward at all. It looks good on paper, but the direct benefit to the reviewer may be limited. So we are left with a strange position: reviewer rewards are restored, but in a context where the publisher may already be receiving institutional payment anyway. One can reasonably ask what, exactly, the reviewer is being given.
That does not make the scheme worthless. It simply shows the limits of publisher-controlled incentives. They can appear, disappear, and reappear in modified form, all without accountability. All pages related to the discontinuation of the original scheme have gone. They deleted them. Considering a journal never disappears, surprising they can delete major policy decisions. Reviewers remain dependent on the goodwill of publishers that have already demonstrated how quickly such benefits can be withdrawn.
Symbolic recognition is not enough
The wider industry still relies heavily on symbolic recognition. Reviewers are offered badges, certificates, ORCID or Publons integration (yay; not), and occasional mentions during Peer Review Week (or the end of year sucker list). These are not useless, but they are not enough.
A badge does not buy out the two to four hours spent carefully reviewing a complex paper. A certificate does not offset the accumulation of requests arriving in the middle of teaching term. An ORCID entry does not compensate for the fact that academic promotion frameworks often barely acknowledge peer review as labour at all.
If publishers want to keep saying that peer review is essential, then they need to behave as if the people doing it matter. That means one of two things. Either they compensate reviewers in a material way, through honoraria, credits with real value, or other meaningful forms of reward, or they dramatically reduce the wider burden surrounding submission and review so that academics are not doing editorial production work for free on top of everything else.
At the moment, too many journals are doing neither.
What needs to change
The first and most obvious reform is simpler, lighter initial submission. Journals should stop demanding publication-ready formatting before a manuscript has cleared editorial triage. If a paper is scientifically sound and within scope, that should matter more than whether every heading and figure label matches a house style. Formatting can come later.
Second, publishers need to reclaim the copyediting and production responsibilities they have gradually pushed back onto authors. If we are paying APCs directly or indirectly through institutional agreements, then researchers should not be spending valuable time on tiny stylistic details that bring no scientific value.
Third, reviewer reward needs to become more serious and more transparent. That could mean direct payment. It could mean credits that are actually portable and useful. It could mean institutional recognition that counts in workload models, appraisal, and promotion. But the current combination of rising expectations and token appreciation is not enough.
Finally, universities and funders need to stop pretending that peer review simply happens by magic. It is labour. If they value it, they should recognise it properly.
Conclusion
Peer review remains essential. But the current publishing model depends too heavily on the assumption that academics will continue to provide endless unpaid labour out of duty, habit, or fear of letting the system collapse.
That assumption is wearing thin.
Researchers are being asked to review more papers, carry more of the editorial burden, navigate more submission bureaucracy, and accept more house-style demands, even where publishers are already being paid through APCs or transformative agreements. JMIR’s return to reviewer credit is welcome, but its earlier absence was poorly explained, the blockchain ambition quietly disappeared, and the value of the revived scheme is less clear in settings where institutional agreements already absorb publication costs. My recent PLOS Digital Health experience, being asked to deal with basic stylistic points such as “Figure” versus “Fig”, only reinforces the sense that too much low-value work is now being passed back to authors.
The problem is no longer just that peer review is unpaid. It is that the entire publication workflow is being rebalanced in the wrong direction. If publishers want rigorous, timely, thoughtful peer review, they need to start giving something back. Not just badges. Not just rhetoric. Not just another dashboard. Something real. Until then, peer review will remain what it increasingly feels like: an essential scholarly function trapped inside an extractive publishing economy.