American Psychological Association
Paper presented to the ICSU Press Workshop, Keble College, Oxford, UK, 31 March to 2 April 1998
After a year of development, the American Psychological Association (APA) launched an all-electronic journal in 1997. TREATMENT, now "Prevention and Treatment," under the editorship of Martin Seligman, PhD began receiving manuscripts in June of 1997, and the first paper was available to readers on the APA web site in September of 1997. Based on reader responses and other external factors, the mission statement and operation was re-conceptualized during the winter of 1998. A revised e-journal format was released in mid-March 1998.
That's what has happened in the last 18 months, the "short story."
Now, for the details. You can't have a truly complete all-electronic journal without establishing an all-electronic editorial office. And, because we are talking about "economics," we need to talk about system design, content selection, technical editing, marketing, and customer services.
The first challenge we faced concerned building the electronic journal infrastructure - and, specifically, the electronic "editorial back office" for the journal. We wanted every aspect of the journal to be electronic, from submission to publication. Therefore, no faxes, no hard disks through the mail, no telephone calls, and no hard copies of manuscripts were welcome. The design needed to have everything done electronically - submission, manuscripts assignment, requests for reviews, action letter correspondence, revisions, re-reviews, technical edits and so forth.
We began building the electronic editorial back office as a "manuscript tracking system" that involved the automatic assignment of manuscript numbers and the permanent tracking of the status of the manuscript. We next added functions for the editor and associate editors to chat among themselves about the manuscript and potential reviewers.
When we got to the point where we wanted to request reviews, we needed to create an accessible, easy-to-use, all-electronic working file of reviewers (including options for adding or deleting reviewers, as well as modifying a reviewer's record). The goal was that when the editor was ready to assign reviewers, he would click on the name of the desired reviewer, and the system would automatically move that person's name and e-mail address into the manuscript file for the specific manuscript - as well as e-mail out the requested review request and the manuscript to the reviewer (and re-filing the manuscript in the "pending review" file).
The system automatically receives back the reviews, thanks the reviewer for the review, informs the editor that the review is in, and files the review in the appropriate electronic manuscript folder (and, when all reviews are in, notifies the members of the editorial team that the manuscript is ready for their final review of the reviews and the manuscript, so an action letter can be sent to the author).
Model editorial action letters were created, so the editor could click on the type of letter he wanted to write, import the author's name and e-address, import the title and number of the manuscript, import the review commentary, and add personalized greetings, additional review commentary from the editor, transitional material and editorial suggestions for changes - and e-mail it to the author. After the action letter is sent, the system automatically files the file folder in the appropriate file - such as, "rejection," "revision pending," or "acceptance" (which results in an automatic hand-off to the technical production staff).
Revised manuscripts go through the same process, until the manuscript is either rejected or accepted. That was the basic design of the electronic journal back office - the essential program.
As we have used the "electronic editorial back office" with other journals, we have had to do varying modifications to address the specific needs of the specific journal operation. Since the basic construction was modular in nature, all modifications have been successfully implemented within 2-6 hours, generally occurring over a 1-3 day period.
This was done over a period of four months at a total cost of approximately $9,000, and modification costs for specific applications have averaged $500, and annual upgrading costs are running $500.
Of course, the system needs to operate on a computer which is physically located somewhere.
For a small society running a single journal and not being particularly concerned about electronic publication and distribution, this could be done on a single PC located just about anywhere (but most likely in the editor's office).
However, because the APA publishes 35 scholarly journals on psychology, and because the APA maintains a large public website as well as vending electronic access to APA PsycINFO® abstract database and the separate APA full-text article database, APA was interested in maintaining all electronic journal electronic back offices on the central APA computer with the editor, associate editors, reviewers, and authors having their appropriate type of access within the system - whether by e-mail, access to secure HTML areas, or transmission of basic word-processing files.
Because of the intent for eventual large-scale electronic distribution, redundancy within the operating system was critical to us. We put in place two separate redundant ISP providers for separate T-1 connectivity to the Internet, so that in the unlikely event that one access provider went down, the other would take over all functions. Similar setups are maintained on the primary and secondary systems to provide the electronic access, with two redundant mirrored servers, redundant shared-disk drives, and so forth. We operate on Sun equipment for electronic publishing, but NT servers with clustering can also be used. To achieve a constant monitoring on the status of the hardware and services running on the primary server in order to perform the changeover to the secondary server, APA uses VERITAS high-availability software, First Watch. Separate licenses are used for each server so that the secondary server can be used for R&D as opposed to idly waiting for the primary server to fail. Netscape Enterprise, which includes secure sockets, is utilized. APA uses PLWEB as its search engine, in customized form. We utilize several security systems.
This is what the editorial back office is sitting on, but as you can see, the described system is really for electronic publishing, electronic access, electronic distribution, and the financial transactions related to all of this. In short, the electronic editorial back office sits in a small corner office of this entire setup, utilizing less than one percent of the overall system capacity.
This is a redundant system designed primarily for the electronic publishing operation for electronic access to the Internet version of PsycINFO® 111-year and Internet access to the APA Full-Text Article database. These initial costs totalled over $235,000 in the first year. Initial hardware cost over $100,000, Internet communications hardware and software cost some $45,000, and initial license on various software totalled $12,000. Less than one percent of these start-up expenses (or about $2,355) is directly attributable to the electronic journal per se. Even the heaviest and most aggressive allocation of cost only increases this number to $6,729 per journal. These numbers are relatively low because of the cost efficiency of the multiple electronic activities in which we are engaged.
Please note that I have not yet addressed personnel costs related to system design, installation, or ongoing MIS technical support.
OK, so now we have an electronic editorial back office and a powerful centralized computer from which it can operate. Now, we need an editor - to run the editorial back office, to recruit and screen, and select appropriate and high-quality content.
New editors require start-up - whether for a print journal or an electronic one, and most of their needs are the same - computer, printer, phone line, software - can cost up to $7500. But, unlike a print journal editor, there is no absolute requirement for a fax machine, copy machine, or file cabinets because there are no paper manuscripts or letters.
As for ongoing operational expenses of an electronic editorial office, there seems to be no getting around the need for some clerical support. For APA published paper journals, that seems to come at a minimum of $3,000 or $105-140 per submission. For an electronic journal, at least an electronic journal such as TREATMENT, the clerical costs are much lower - more on the order of about $1,000. This is independent of submissions, as the major tasks revolve around set up and maintenance of the reviewer pool.
I have already described the program for the electronic back office, so we have really walked through the operation of the editorial back office and the basic content review of the submitted material. Key to repeat and emphasize is that the author submits directly on-line their full-text manuscript and that all aspects of the manuscript handling and reviewing is done electronically. The end of the editorial process, for those manuscripts that are accepted, is a hand-off to the technical production office.
The APA journal's technical production office does a standard technical edit on manuscripts intended for the electronic journal, with proposed editorial changes as well as author queries, which are electronically sent to the author for review, approval/rejection (or modification), and answers to queries. I tried hard for a "no technical edit" approach, but it simply proved impossible. All authors, even the most accomplished, need copy editing of some sort. Once the final content form is available, coding for Internet display is done. We primarily operate the APA website using SGML coding, which is transformed on the fly into HTML for display purposes. A small percentage of material is directly HTML coded, but we are working to eliminate this.
When a manuscript is finally accepted, it ends up (after some intermediate steps I will go into shortly) on the APA TREATMENT website.
The e-journal is not limited to just being an electronic version of what we publish in print. We have tried "live" electronic chats with article authors at a set time and day. We found little interest in this (at least, as yet). We provide for non-reviewed, "bulletin board" discussion and chat about articles - which we do not plan to archive (and would rather just have it up for the first 3 to 6 months). We also have a mechanism for fully peer-reviewed critique and commentary. This might be done in the first week, eight months later, or 20 years later - but it is directly linked to the target article, so one does not have to discover it and locate a copy from some later issues, as is the case with print journals.
Since we are currently offering the journal free, we are not actively marketing it. However, in thinking ahead to when it is a subscription product, we have considered the following -
Electronic Marketing Avenues
And, conventional Marketing Avenues
We can envision marketing plans that range from NO direct costs to ones that cost upwards of $25,00 (plus staffing expenses).
Accounting and bookkeeping - $0 to $25,000
Customer Service: 14-16 hours per week for 5,000 subscribers, at least some hours every day of the week - $20,000 (includes benefits and overhead)
In between the acceptance of a manuscript for publication and the actual appearance of the online journal, there are a number of steps that tend to be overlooked or minimized when we talk about electronic journals: technicians, programmers, designers; technical editors, and marketers. Even after the e-journal material is selected and posted, there is maintenance, trouble-shooting, and technical support that needs to be provided - more skilled people tasks. With print products, publishers' involvement with content and product more or less ended when we dropped the printed issues in the mail, but with electronic publications, the involvement of the publisher potentially continues.
With electronic publications, there are more "behind the scenes" tasks than ever before, and these tasks require skilled people. And, as all publishers know, skilled people usually come with a price tag (unless, of course, you can get someone to do it for free - a more and more difficult task to accomplish of late).
What we do when we go from paper to electronic is trade one set of costs for another - there are NO typesetting costs, or outlays for paper, print, postage. But, there are NEW needs for hardware, software, technical support, and system maintenance.
If you are in a university setting and can get some equipment, service, and skilled staff for free - wonderful. If you can get grant money, great. Otherwise, savings tend to be offset by new sets of expenses.
New law of conservation:
For every innovation that results in lower costs in one area of publishing, there is an expense that will equal the savings.
The REAL savings is in time - 2 days to 2 weeks after acceptance to distribution vs the usual 6 months.
The real VALUE is in added features:
There are a lot of new electronic journals, but they are still in the infancy stage of development. Authors have yet to accept the electronic-only journal as being a competitive outlet for their best work. Readers are not generally willing to subscribe to them. The full potential of what they can offer has yet to be realized. But, it's only a matter of time.
Authors submit directly online and do a majority of the data entry for information conventionally handled by an editorial assistant.
[TREATMENT Submission Page - #2]
And, when the manuscript is received, it is handled in the "Journal Back Office" -
[Journal Back Office #3, #4, and #5]
(Demonstration of manuscript submission, manuscript assignment to reviewers, manuscript tracking, etc. in an electronic journal office.)
Returning to the economics of electronic journals -
In between the acceptance of a manuscript for publication and the actual appearance of the online journal, there are a number of steps that tend to be overlooked or minimized when we talk about electronic journals:
These tasks require skilled people, and as all publishers know, skilled people usually come with a price tag (unless of course you can get someone to do it for free - a harder and harder thing to do these days).