SSRN Responds to My Post

Last week, I blogged about my primary frustration with SSRN: the long delay that occurs between uploading a revision to an essay and it replacing the old one. Gregg Gordon, the President and CEO of SSRN, graciously replied to my post — and then to my follow-up questions. Gregg has given me permission to reproduce our exchange, so here is his initial reply (edited for readability):

Professor Heller,

Thank you for the invitation to respond to your blog post of 3/26/07. You are correct SSRN spends a significant amount of time reviewing submissions to the SSRN eLibrary database and I would very much like to eliminate that use of staff time. Unfortunately, based on our experience, we find the cost of our efforts to be very beneficial. A higher than expected percentage of submissions have errors including the wrong PDF uploaded, not all of the authors listed, wrong authors being listed, improper formatting, serious typos in the title or abstract body, etc. Since our users expect high quality and we are very concerned about not embarrassing an author, we have chosen to incur the cost of reviewing all submissions for completeness.

As you may know, we give submitters the option to make their new submission “Immediately Available” with a watermark. We have not developed this option for Revisions because the revision process is different from the new submission process (depending on the complexity and type of changes made) and requires similar but not exactly the same functionality. With our limited development budget, we are only making minor changes to the existing submissions/revision processes so that we can complete our new Simple Submission system. This new system, currently in the alpha-stage, will allow for Revisions to be “Immediately Available” and should greatly improve efficiency for both submitters and SSRN staff.

In general, our review of a Revision should always take less than 24 hours. From late February through mid-March, we receive thousands of submissions for the EFA and AFA annual meetings. These additional submissions delay our ability to process all submissions. Your revision of “Retreat from Nuremberg” was treated as a new submissions because the new APS version is overwriting the previous WPS version. If you experience delays of more than 24 hours going forward, please let me know and I will check on it.

Please let me know if I have addressed your concern and if you have any further questions.

Thanks again,
Gregg Gordon

I responded by thanking him for the e-mail and expressing my belief that the “immediately available” option for revisions will be great. Unlike some others, I don’t particularly care about the watermarks; I just want scholars to be able to immediately download my essays. I also made a few additional comments, to which he quickly replied (my comments are in bold):

First, although your argument in favor of review makes sense, I don’t find it compelling. Why should I, someone who spends a great deal of time ensuring that my SSRN submissions are accurate, have to suffer because others are less meticulous? Even if a majority of submissions have problems, that doesn’t justify subjecting me to week-long delays. Moreover, your review system encourages sloppiness — scholars know that if there is a major problem, you will fix it for them. I would happily waive your review process if given the opportunity — if something turns out to be wrong, that’s my fault, not yours.

I appreciate your point of view and agree completely that a week is too long to wait. We have grown tremendously over the past few years (annual submissions now exceed 30,000 and annual PDF downloads are approaching 4 million) and this growth has caused some delays. I receive a report every day with the number of items submitted each day for the past week and how long it takes to review them. This issue is a priority for us.

I also realize that it is possible that at some level we may “encourage” author sloppiness. We have 73,000 authors in the SSRN eLibrary and receive submissions from a variety of sources, including co-authors, publishers, XML feeds, working paper coordinators, administrative assistants, etc. and despite everyone’s best intentions mistakes are made.

When an assistant makes a mistake and is listed as the first author, above a prominent scholar, it causes a lot of frustration. In our opinion that frustration is avoidable. I regularly receive emails from authors thanking us for correcting their errors. I also receive emails from authors asking why the titles/abstracts of their submissions have errors and how it happened. I rarely receive emails from users complaining about the quality of the content on SSRN. They may not like or agree with content but they do not complain about errors. In my experience, when someone finds an error, they do not blame the author, they generally blame SSRN.

You are correct that their currently is a “time” penalty for those that submit perfectly. Our solution for them is to opt-out of the review process by selecting “Immediately Available” when submitting. Once we complete our new Simple Submission system, I think the time penalty will be significantly reduced and we will allow Revisions to be “Immediately Available” too.

Second, you say that your revision review process should always take less than 24 hours. In the three years that I have used the system, with at least 15 revisions, it has never taken less than three days. And I’ve been keeping track!

I apologize for that level of delay. Please email me next time you make a Revision and let me know how long it takes so that I can track it internally.

Third, I don’t understand why you would treat an WPS-APS change as a new submission, unless you verify acceptance with the journal — which I seriously doubt. The only changes I made were WPS-APS, the name of the journal, and the file. That doesn’t seem like very much — certainly not enough to justify the eight days it took to approve the revised essay.

The eight days was due to the very large inflow from two conferences but it is unacceptable. We do not verify acceptance with an APS journal. Unfortunately, switching from WPS to APS is a legacy process that still requires some manual review. The legacy issues relate to the eJournal classifications (in many networks we have both a WPS and APS version of the eJournal) and confirming the paper is properly classified in each of the eJournals. We are also looking into modifying our paper types to better match the changing environment and eliminate this problem.

I’m not completely convinced, but I very much appreciate Gregg taking the time to respond to my post and follow-up comments. It’s clear that SSRN cares a great deal about delivering a quality service.

Any additional thoughts from our readers?