The Decision Letter, Part II

23 September 2019, 1217 EDT

The basic principles that should guide letters and their RFDs hold across every kind of decisions. However, we need to recognize important differences between, say, a rejection and an R&R. In this post, I lay out my thoughts about letters for the types of decisions that we made at ISQ. Not all journals use the same categories or mean the same thing. For instance, a (rare) “reject and resubmit” at ISQ meant that we would consider a revised version of the paper as an entirely new submission; at some journals, a “reject and resubmit” is equivalent to “major revisions” R&R.

Desk Rejections

Given the volume of submissions journals now receive, many editors use desk rejection to winnow the manuscript pool. This, in turn, tends to generate controversy—and not without reason. Good arguments exist both for and against desk rejecting manuscripts that editors judge are highly unlike to survive peer review.

However, most desk rejections happen for reasons that aren’t very controversial. Journals receive plenty of submissions that fall outside of their scope, don’t meet their basic standards on matters like citations or language, and otherwise simply can’t be sent out for peer review. I used to call these “procedural” (as opposed to “substantive”) desk rejections.

It might be tempting to use an unmodified form letter for procedural desk rejections. But I think even procedural desk rejections require a minimally tailored RFD. If a piece is outside of a journal’s scope, the letter should describe the reasons why. If a piece doesn’t meet standards for appropriate citations, the letter should provide examples of its problematic practices , as well as point the author toward relevant guidelines.

Substantive desk rejections are a different animal. They demand a more extensive letter. Editors desk reject otherwise reviewable papers because they’re trying to a) reduce the burden on their referee pool and b) spare authors waiting thirty or more days to receive a rejection. Those are both good reasons to issue substantive desk rejections. But those desk rejections deprive authors of feedback from referees. Part of the purpose of the peer-review process is to provide such feedback, with the idea that it will at least help authors secure publication at other journals.

Thus, editors have an obligation to write an RFD that provides at least a partial substitute for those absent referee reports. That means the RFD should include developed warrants and, ideally, suggestions for improving the paper.

There’s another reason to provide some non-trivial level of detail in the RFD for a substantive desk rejection. The content of such an RFD provides authors with grounds for appealing decisions. If editors simply send along form letters with no real content, then authors have no way to seek recourse for bad editorial decisions. Put differently, the warrants of the RFD create some kind of “check” against editorial mistakes.

(Yes, I do think that journals should have standing procedures for appeals. I suppose this implies that I might need to write a future post about appeals.)

Rejections

Editors of established journals write many rejections. For better or worse, they tend to develop specific triage procedures for rejection letters.

What does this mean in practice? Many editors, I assume, divide manuscripts into categories based on their initial interpretation of referee reports. Of these categories, two or three place the submission on track for a potential rejection. For example, these might include:

  • “Easy” rejections, in which all of the reviewers recommend rejection and their reports appear superficially consistent with what the editors view as reasons for rejection.
  • “Likely” rejections, in which the reports are split between rejections and revise-and-resubmits, but the contents of the reports seem to suggest the manuscript might be borderline.
  • “Possible” rejections, in which the net impression given by the reports are ambiguous.

I’ll be honest here: I don’t think many editors give a lot of attention to manuscripts in the “easy” category. At high-volume journals, the manuscript may not get much of a second glance. After all, at many journals at least one editor already read the manuscript when it initially came in. Instead, there’s a good chance the editors will simply generate a decision letter by aggregating and synthesizing the referee reports. This is pretty easy to do. In these cases, there are almost always similar issues flagged by the reports. Editors can easily restate those concerns as warrants in the RFD. One assumption behind this kind of triage is that if multiple referees converge on serious concerns, then then those concerns are probably correct.

Keep in mind that editors also presume, with good reason, that the odds do not favor a paper overcoming multiple negative reviews via revisions. So even if you know that there’s some chance that one or two of those warrants reflect referee mistakes… well, there’s a stack of other manuscripts that need your attention, and it’s not getting any smaller. Editors play an “attention conservation” game, and pieces with ‘clear rejection’ reports just aren’t going to get the same care as those with more borderline reports. Should they? Probably. But that requires either a very large editor-to-submission ratio or much more extensive course releases than most academic editors get.

The problem with all of this is straightforward. Two or three anonymous reviews can all deliver misguided reports. They can even make similar errors. Yes, it’s easy to argue that if the referees are all getting the same things wrong, then the problem likely lies in the style and presentation of the manuscript. Sure. Probably. But, ideally, editors ought to at least ‘cross check’ the major objections against the paper. It’s just hard to fault overburdened editors for cutting corners.

I’m sure you’ve figured out by now that all of this is another reason to provide clear and precise warrants in the RFD. If the referees really messed up and the editors didn’t catch it—either because they triaged the manuscript or because they missed something—then this gives authors an opportunity to appeal the decision. At the least, they can send a polite letter expressing concern.

(I should note here that the “polite letter expressing concern” is sometimes the best opening move in the pursuit of an appeal. It certainly can make editors more disposed to give you a fair initial hearing, and if they agree with you then they might move the ball forward on your behalf.)

Clearly specifying the warrants for a rejection—and, if appropriate, placing them into a contextual hierarchy of importance—has another benefit. It signals to authors and referees which aspects of the reports mattered, and to what degree, to the editors.

I’ve already mentioned that I now think RFDs need to be explicit about what didn’t drive a decision. When I talk about this issue, I usually mention how ISQ once received an incredibly professional, thoughtful, and persuasive appeal. It focused entirely on one of the referee reports. We agreed that report was very low quality, which is why our RFD largely ignored it. We mentioned only one of its concerns, and that concern was shared across all of the reports.

It isn’t just that we could have saved the author time by being more explicit. It’s also that doing so would have made clearer to the referee that their report was not up to ISQ standards. Part of an editor’s job is to nudge academics to be better referees.

Calibrating Tone in Rejection Letters

Even in apparently straightforward rejections, it’s generally best practice to identify some positive aspects of the manuscript and to provide some of the warrants in the form of suggestions. As I noted in my prior post, this means phrasing warrants along the lines of, say, ‘the manuscripts needs to do a better job of answering [some objections] and perhaps, as R1 suggests…’ rather than, say, ‘the referees find the manuscript’s handling of [some objections] totally flawed.’ It can also mean talking about the referees being unconvinced, rather than assigning “right” and “wrong” to the referees and the manuscript.

The difficulty here is that if editors overdo this then it can obscure the degree that the paper suffers from real problems. Authors may find it hard to understand why they received a rejection rather than a revise-and-resubmit invitation—especially since most manuscripts are rejected, in part, based on the scope and scale of needed revisions. So the RFD on a clearcut rejection requires a balancing act. This is even more the case with the RFD for more borderline rejections. If it reads too much like an R&R, the editors can undercut their own explanations.

Despite all of this, I often found it useful to write the first draft of a ‘hard call’ as an invitation to revise and resubmit. Once I laid out the case for an R&R on paper, I could then see if it made sense: if the nature of revisions really fit with what we associated with revise-and-resubmit outcomes. I’d also like to think that this process produces more palatable, and helpful, rejection letters. However, it does require deliberately shifting some of the language in more of a negative direction when changing a revise-and-resubmit letter into a rejection.

Rejection After Review

Rejections are unpleasant for both authors and editors. I don’t know any editors who enjoy rejecting manuscripts, even though it does mean one less piece in the queue. But the second worst letters to write—the worst deal with academic misconduct—are rejections after review.

I think that most editors dislike declining a manuscript that’s been through one or more rounds of revision. Editors usually invest a lot of energy in a revise-and-resubmit invitation. They engage closely with the ideas, arguments, and limitations of those manuscripts. Editors want them to succeed.

Moreover, the vast majority of ‘failed revisions’ come from manuscripts that the editors made a conscious decision to take a gamble on. That is, they offered an R&R despite some pessimism about the likelihood of eventual acceptance—they decided the potential reward was worth the risk. Thus, when authors fail to convert an R&R it may suggest that editors erred in their initial decision.

As unpleasant as it is for an editor, unsuccessful revisions are worse for authors. They often feel like wasted investments, and they always represent dashed hopes. Because peer review is arbitrary and capricious, there’s no guarantee that a failed R&R will actually make it easier to land the article elsewhere. So it’s just bad news all around.

When we started out, we (stupidly) did not have a specific template for reject-after-revisions. We created one, complete with boilerplate explaining how hard such decisions are, especially for authors. As I’ve stressed before, such boilerplate is actually sincere. Regardless, editors should approach reject-after-revisions with particular care. Not only because of the considerations that I’ve mentioned, but because they are the most likely to be appealed (and successfully). So the RFD needs to be crystal clear, both to justify the decision and to facilitate a possible appeal.

(Indeed, at least one of the most well-received manuscripts we accepted while I was editing ISQ was initially rejected after second-round review!)

The Revise and Resubmit

Rejection letters should provide feedback that helps authors secure publication elsewhere. But he aforementioned vagaries of the peer-review process mean that the feedback authors receive in a rejection may not provide much help in getting the manuscript accepted at a different journals. This is not the case for revise-and-resubmit letters. In revise-and-resubmit invitations, the editors, drawing on the referee reports, set parameters for what changes a manuscript needs before it can be accepted.

Journal editors play two major roles. First, they are gatekeepers for their journal—and, hence, to various degrees, relevant fields. Second, they are editors in the dictionary sense: persons who are “in charge of and determines the final content of a text”. When editors reject a manuscript, they primarily occupy the first role. When editors offer a revise-and-resubmit invitation, the second role kicks in.

In my view, that means that editors should craft their revise-and-resubmit letters with the objective of helping authors secure publication. Full stop.

For an R&R invitation to do its job, editors need to very clearly communicate the “hierarchy of warrants” that I discussed in the previous post. If editors fail to do this, then authors lack key guidance for how to handle the disparate suggestions made by referees. Good revise-and-resubmit letters also signal priorities to referees in the next stage of review. This ranges from sorting out what the editors consider important—and to what degree—to providing an explanation for why the editors ‘overruled’ acceptances and rejections.

Some editors—mostly because of time pressure, I imagine—basically abdicate this crucial editorial role. At the extreme, they slap a form letter on the referee reviews and call it a day. I’ve seen situations where referees advocate incompatible directions for revision and the author receives nothing more than generic instructions.

I’m only engaging in moderate hyperbole in calling this editorial malpractice. Writing confusing decision letters, getting things wrong, or not realizing that referees contradict one another? These kinds of things happen to even conscientious editorial teams. To repeat myself, some rate of error is inevitable. But there’s no excuse for generic R&R letters with no, or the barest of, tailoring to the content of the reports.

When it comes to significant contradictions among the referee reports, revise-and-resubmit letters need to implicitly or explicitly deal with the problem. They can do so implicitly by constructing a hierarchy of issues that makes clear that these contradictions don’t involve anything dispositive. They can also acknowledge them and at least try to provide the author some guidance for how to proceed. Either way, editors should try to make it clear that they understand whats going on when authors face a dilemma in addressing referee reports, even if a particular letter doesn’t provide definitive guidance for how to proceed.

But, really, it’s the job of a revise-and-resubmit letter to help authors make those decisions. Sometimes this means telling authors what direction the editorial team wants them to take. Other times, this means working through the possible implications of a choice, but making it clear that the decision is in the hands of authors.

One might argue that editors shouldn’t effectively second-guess referees, who generally are more qualified to evaluate a specific manuscript than they are. But, for better or worse, ultimate authority in the process lies with editors; they make the final decision about whether to accept or reject a manuscript. So they need to act accordingly.

Of course, the revise-and-resubmit process changes the game in important ways. It effectively transforms the peer-review process into a three-way dialogue among the editors, referees, and authors. Revision memos—both confidential ones to the editors and ones addressed to both the authors and editors—play a major role in this shift. I’m not sure authors always realize this, or recognize just how important memos are. But that might be a subject for a future post.

One last thing about the R&R process. These features—that the editors should be doing their best to help the manuscript get published and that the process becomes more of a three-way dialogue—have their limits. While we did answer questions when authors emailed us, we exercised caution about moving beyond clearing up ambiguities or expanding on matters already in the letter. We refused to read outlines or otherwise do things that would make ‘willingness to engage the editors’ a source of unfairness across submissions.

(I also learned the hard way to avoid, except in unusual circumstances, talking about manuscripts with authors over the phone. Everyone needs a written record of correspondence as a check against misunderstandings.)

Conditional Acceptances and Acceptances

There’s not a lot to say about these kinds of letters. The main thing is that a conditional acceptance letter needs to be clear about what a conditional acceptance means. Not all journals issue conditional acceptances; the stakes of this kind of decision varies not just across, but also within, journals. A conditional acceptance could mean that, for example, the editors:

  • Expect to accept the manuscript once it satisfies the conditions;
  • Are offering an R&R but without further external review—that is, the editors might wind up declining the manuscript; or
  • Intend to send the manuscript back out for review but will accept the manuscript if it satisfies the few remaining concerns.

I’ve never actually seen a conditional acceptance that didn’t make clear what its ‘terms’ were. But, you know, never say never.

Final Thoughts

One of the most important thing for people to realize about decision letters is that even ‘pretty good’ journals are increasingly overwhelmed by the number of submissions that they receive. No matter the ‘type’ of decision that they are writing, editors are always mindful of a stack of other manuscripts that need attention of one kind or another.

This situation is exacerbated in fields like International Relations, where field-wide journals consider manuscripts on a wide number of subjects that use very different methodologies. While many journals have responded to the “numbers + heterogeneity” problem by expanding their editorial teams, it’s inevitable that editors will wind up shifting modes across submissions and writing letters that require different kinds of criteria. So there’s a lot of room for things to go sideways.

The other is that decisions aren’t personal. Most editors, I hope, are focused on manuscripts—they couldn’t care less about the identity of authors. In most circumstances, if they do care about the identity of authors, they’re doing it wrong. If editors sometimes seem overly bureaucratic and procedural, that’s because doing so makes their lives easier and establishes checks against bias. Journals, moreover, should have procedures in place to limit conflicts of interest. Most of the major IR journals implement strict firewalls to reduce editorial bias. Ideally, editors would use triple-blind review, but that’s not always compatible with a smoothly functioning journal.

Anyway, thoughts? Questions? I hope you all are finding this series useful.