Not that I'm defending the current, broken system becuase it really doesn't provide either of these benefits, but the idea is that the cost of (coordinating) peer-review (as the reviewers are often unpaid) and actual editing cost money. If services like that were actually provided, some fee would be understandable.
One of the reasons the figures end up higher than you may expect is that only the successfully published papers are paid for (usually). So depending on the rejection rate, which I think is ~95% in the case of Nature, you're paying for those other 19 papers to be checked.
PLOS for example is non-profit and charges up to $2900.
> So depending on the rejection rate, which I think is ~95% in the case of Nature, you're paying for those other 19 papers to be checked.
Yes, the other 95% need to be checked, but reviewers are recruited from academics and typically do this task for free. The contribution that the publisher does at the stage of manuscript evaluation is marginal.
It'll go through the editors first, in the case of Nature these are paid positions. There's a paid "Associate Editor" position at PLOS too, responsible for "Assessing new submissions and guiding manuscripts through the review process" and more.
A submitted paper does not simply turn up in reviewers inboxes, there are steps in between.
I think people pick a few elements of the whole process and then say everything else is negligible. If paid, the academic editors and reviewers may well end up costing a huge amount compared to the spend elsewhere, but that's not the same thing as saying those other costs are small. The proportion here is largely irrelevant.
The associate editor position on glassdoor is about £40k/year, which is £45k including tax costs. Let's say that's £47k including pension contributions as it works out neatly. In the UK there are 47 working weeks, roughly so that's £1k employment cost per week purely on that one employee. That's £25/hour. At an 80% rejection rate that's actually £125/hour, at 90% it's £250/hour and 95% that's £500/hour on accepted papers (not quite, but useful for the comparison). For only a single employee, and only their direct salary.
Of course now we need to add things like the HR costs, hiring costs, building rent, computer equipment, management, etc. Double? How much time of their day actually goes to the core task and not other meetings/etc. All these things multiply up and I'm really not that surprised that the costs go up to these amounts.
I have absolutely no doubt that if the other editors were paid and the reviewers were paid then this would go up dramatically, but that's a different issue.
What exactly do the editors do? I have published more than a dozen papers across different publications like ACM and Springer. I edit the article, I am the one who provides in the format as the conferences/journals - the journals just "print". The reviewers don't charge - I know, I have been a reviewer for Elsevier too. What exactly do the editors do?
> I think is ~95% in the case of Nature, you're paying for those other 19 papers to be checked.
All those 95% of papers checked are done by reviewers for free. I don't see any reviewers being paid. I know because I have reviewed papers across multiple conferences and journals.
The free reviewer work is not really relevant for this calculation though, unless there is no work done by the journal for those papers. Since papers are not submitted directly to the peer reviewers, there is at least some work happening.
Free peer review mostly only tells us that the total cost could go up significantly if they were paid.
Maybe. As someone who has done quite a bit of editing of relatively technical material and coordinated review and rework cycles, I wouldn't be surprised if you're looking at a couple of days of work for a paper. Certainly a lot more than for a quick copyedit of an online article. So you're into the more than $1K range pretty quickly.
$5k though? That still seems a bit much.