An alternative model for peer review

There is no doubt that the current model of peer review is an effective but inefficient system. The high quality of publications that complete peer review is a testomy to the effectivity of the peer review system, as poor papers rarely get accepted in well reviewed journals. However the efficiency of the review system is very low.

Consider that the highest ranked journals have acceptance rates of around 10% and even the middle-ranked journals have acceptance rates of less than 50%. Most papers get published sooner or later, but with the career reward of publishing in high impact factor journals, it is not unusual for a publication to get rejected four or five times as the authors work their way down the journal ranking list. Considering that each review will generally consist of three reviewers, a single paper that had a tough time could consume the (unpaid) time of fifteen reviewers before it is finally accepted. This is an enormous burden on the scientific community, and is largely a wasted burden - afterall, each journal editor only gets to see three of those fifteen reviews when making a decision to accept or decline an article. It also considerably slows down the dissemination of information, as it is not unusual for the entire review process to consume a year or more.

So let's consider an alternative model for peer review, one which keeps the critical aspects that provide effectiveness, but which changes the policies that produce inefficiency. Consider now a consortium of four or five publishers, which may include 20 journals that publish papers on immunology. Rather than authors submit to the individual journals, the authors would submit to a centralised editorial staff, which is paid for by the publishers but which is independent of each journal. An immediate advantage would be the ability to have many more specialised editors available, allowing for better decisions on choosing and assessing the reviews.

Each paper would then be sent out to five or six reviewers, and the reviews would be made available to each of the journals. The editorial staff at the journals would be able to make an assessment of the paper and put forward an option to accept, conditionally accept or decline the paper. This information would be transmitted back to the consortium, and would be provided to the authors. The authors would then be able to make their choice of which offer to accept. In effect, each journal would be making a blind offer to the authors to publish their paper, with full knowledge of the reviews but without the knowledge of whether the other journals put in a bid.

Consider the benefits of this alternative model to each player:

1. The journal gets to judge on more complete information, with double the number of reviews available for each paper, selected by more specialised editorial staff.

2. The reviewing community will more than halve the number of reviews required, while actually providing more information to the journals.

3. The authors will no longer have to make strategic decisions in choosing where to submit, they will simply submit to the consortium and have the option to publish in the top ranked journal which is interested in the paper.

4. The scientific community will have access to cutting-edge research months or even years earlier than under the current system.