Robert Kelchen

Some Thoughts on the Academic Peer Review Process

Like most research-intensive faculty members, I receive regular requests to review papers for legitimate scholarly journals. (My spam e-mail folder is also full of requests to join editorial boards for phony journals, but that’s another topic for another day.) Earlier this week, I was working on reviewing a paper submitted to The Review of Higher Education, one of the three main higher education field journals in the United States (Journal of Higher Education and Research in Higher Education are the other two). I went to check one of the submission guidelines on the journal’s website and was surprised to see that the journal is temporarily closed for new manuscript submissions to help clear a backlog of submissions.

After I shared news of the journal’s decision on Twitter, I received a response from one of the associate editors of the journal. Her statement astonished me:

This sets off all kinds of alarms. How can a well-respected journal struggle so much to get qualified reviewers, pushing the length of the initial peer review process to six months or beyond? As someone who both submits to and reviews for a wide range of journals, here are some of my thoughts on how to potentially streamline the academic peer review process.

(1) Editors should ‘desk reject’ a higher percentage of submissions. Since it can be difficult to find qualified reviewers and most respectable journals accept less than 20% of all submissions, there is no reason to send all papers out to multiple external reviewers. If a member of the editorial board glances through the paper and can easily determine that it has a very low chance of publication, the paper should be immediately ‘desk rejected’ and quickly returned to the author with a brief note about why it was not sent out for full review. Journals in some fields, such as economics, already do this and it is sorely needed in education to help manage workloads. It is also humane to authors, as they are not waiting several months to hear back on a paper that will end up being rejected. I have been desk rejected several times during my career, and it allowed me to keep moving papers through the publication pipeline as a tenure-track faculty member.

(2) Journals should consider rejecting submissions from serial free riders. The typical academic paper is reviewed by two or three external scholars in the peer review process, with more people potentially getting involved if the paper goes through multiple revise and resubmit rounds. This means that for every sole-authored paper that someone submits, that person should be prepared to review two or three other papers in order to maintain balance. But in practice, since journals prefer reviewers with doctoral degrees and graduate students need to submit papers in order to be eligible for academic jobs, those of us with doctoral degrees should probably plan on reviewing 3-4 papers for each sole-authored paper we submit. (Divide that number accordingly based on the number of co-authors on your submissions.) It’s okay to decline review invitations if the paper is outside your scope of knowledge, but otherwise scholars need to accept most invitations. Declining because we are too busy doing our own research—and thus further jamming the publication pipeline—is not acceptable, particularly for tenured faculty members. If journals publicly commit to rejecting submissions from serial free riders, there may be fewer difficulties finding reviewers.

(3) There needs to be some incentive for reviewers to submit in a timely manner. Right now, journals can only beg and plead to get reviewers to submit their reviews within a reasonable time period (usually 3-6 weeks). But in my conversations with journal editors, reviewers often fail to meet that timeline. In an ideal world, journal reviewers would actually get paid for their work like many foundations and scholarly presses pay a few hundred dollars for thorough reviews. Absent that incentive, it may be worth establishing some sort of priority structure that rewards those who review quickly with quick reviews on their own submissions.

(4) In some cases, there needs to be better vetting of reviews before they are sent to authors. Most reputable academic journals have relatively few problems with this, as this is the job of the editorial board. Reviews generally come with a letter from the editor explaining discrepancies among reviewers and which comments can potentially be ignored. But the peer review process at academic conferences has more quality control issues, potentially due to the large number of reviews that are requested (ten 2,000-2,500 word proposals is not uncommon). It seems like reviewers rush through these proposals and often lack knowledge in the subject matter. Limiting the number of submissions that any individual can make and carefully vetting conference reviewers could help with this concern.

By helping to restrict the number of items that go out for peer review and providing incentives for people to fulfill their professional reviewing obligations, it should be possible to bring the peer review timeline down to a more humane 2-3 months rather than the 4-8 months that seems to be the norm in much of education. This is crucial for junior scholars trying to meet tenure requirements, but it will also help get peer-reviewed research out to the public and policymakers more quickly. Journals such as AERA Open, Educational Evaluation and Policy Analysis, and Economics of Education Review are models in quick and thorough peer review processes that the rest of the field can emulate.