Hindawi Publishing: A Working OA Model
by Nancy Davis Kho
In the 3 years since academic publisher Hindawi Publishing Corp. moved to a strictly open access (OA) publishing model, the perception of OA has shifted. Once viewed as a radical upending of traditional subscription models designed to move academic research more quickly and cheaply into the hands of readers, OA has edged closer to mainstream acceptance by publishers and authors. Peter Suber, who publishes the SPARC Open AccessNewsletter, says, “OA journal publishers were once peripheral; now they’re a sizable minority.” While controversy remains around OA’s long-term implications for scholarship and publishing, Hindawi’s recent growth and strategic initiatives show that OA is here to stay.
Hindawi opened its doors in 1997 as a commercial publisher of peer-reviewed journals, covering a range of academic disciplines under a traditional subscription model. Founders Ahmed Hindawi and Nagwa Abdel-Mottaleb decided to set up company headquarters in Cairo in part because it would let them draw on a population of highly skilled employees at a relatively low cost and enable them to keep all production services for their journals in-house.
By 2004, the company began experimenting with a hybrid OA model that enabled journals to include subscription and OA content. Emboldened by the success of the hybrid model, Hindawi moved to an all-OA approach for its journals by 2007. By the end of that year, the company announced a joint initiative with SAGE Publications to publish a collection of OA journals, leveraging Hindawi’s production technology and SAGE’s editorial expertise.
Keys to Rapid Growth
The move to an all-OA model has been paying off. Paul Peters, Hindawi’s head of business development, says the company now publishes 275 journals, about 10% of which are carried on the SAGE-Hindawi Access to Research platform.
Reflecting the growth in the number of journals published is a commensurate jump in submissions. “We have received 21,000 submissions over the past 12 months and are now averaging about 2,000 a month,” says Peters. “Of those, we’ve published 7,500 articles in 2010, compared to 3,650 articles last year.” The most widely read of the Hindawi journals, including the Journal of Biomedicine and Biotechnology and the EURA SIP Journal on Wireless Communications and Net work ing, receive nearly 1,000 submissions each year.
Peters sees Hindawi’s stringent peer-review process and a rejection rate that hovers at about 60%–65% as proof of OA’s long-term viability. “When OA was getting started, there were two camps: publishers with a long-term vision for OA publishing, like BioMed Central, Public Library of Science [PLoS], and Hindawi, and those who were in it to make a quick buck,” he says. “Authors were being bombarded with offers to publish, but over time, the second group of publishers got a bad reputation for quality. Luckily, they seem to be fading into the background.”
The importance of establishing industry standards for high-quality OA peer review led Hindawi to play a founding role in establishing the Open Access Scholarly Publishers Association (OASPA; www.oaspa.org). According to OASPA, the group’s aim is to support and “represent the interests of Open Access (OA) journal publishers globally in all scientific, technical, and scholarly disciplines … through exchanging information, setting standards, advancing models, advocacy, education, and the promotion of innovation.”
Who’s the Customer?
In traditional scholarly publishing, the customer has been a library subscriber to journals. “We were terrible at dealing with libraries,” says Peters. “Being in Egypt, you just can’t send a sales team over to talk to them about our journal offerings.”
But he says Hindawi retooled its focus on who the paying customers were and designed its processes to serve them. Authors of the articles published in Hindawi journals pay an article processing charge (APC) for their research to be distributed freely; the APC ranges from zero for publication in newer journals to $1,500 for the most popular publications.
“We really focus on trying to provide the best possible author services at Hindawi,” says Peters. Because authors are motivated to have their research distributed as quickly as possible, the company has made the speed of publication a key metric. “We’ve reduced the average amount of time from acceptance to final publication from 75 to 35 days,” says Peters. “Our goal is to have it under a month by the end of 2010.”
Another factor that appeals to authors is Hindawi’s in-house production services, teamed by a staff of 120. “For instance, we are starting to publish figures and mathematical equations as scalable vector graphics for the online version of our articles, which enables us to maintain a consistent look across different browsers and screen sizes,” says Peters.
Valid Concerns About OA
In the world of STM publishing, the idea of shifting the cost from subscribers to researchers under the OA model continues to provoke controversy. “There are certainly some valid concerns around the relative publication costs to the cost of performing the research itself,” says Peters. “In well-funded fields like biotech, chemistry, and engineering, the APC is a tiny percentage of the overall research cost,” what Suber characterizes as “vanishingly small, considering the average National Institutes of Health … research grant is $700,000.” But in disciplines such as mathematics and social science (and Hindawi is just starting to move into the latter), the APC may be prohibitive. And unlike BioMed Central, Hindawi does not offer reduced APC pricing to researchers in emerging markets who may be even less able to afford it. Matthew Cockerill, managing director of BioMed Central, says, “Access to information is a real problem in developing countries, but the rapid spread of low cost digital technology, together with open access, is really helping to address that.” He adds, “We needed to ensure that publication fees would not be a barrier for these researchers, hence the automatic waiver scheme.”
Suber makes the point that Hindawi’s article processing charges, which top out at $1,500, tend to be a bit lower than those of BioMed Central. “And it’s not true that the only choice is between high and low APCs,” he says. “Seventy percent of OA journals charge no fees of any kind.” In Hindawi’s case, about 10% of the journals carry no fee for article processing; BioMed Central also charges nothing for submissions to 10 of its publications.
One way that Hindawi is tackling the issue is by introducing institutional memberships. The program is based on a single flat-rate payment that covers all accepted articles from the organizations that are member institutes. It’s calculated based on the total number of submissions from the previous year; once the institute joins, authors who submit during the 12-month period know that they will not be charged at any point, and administrative overhead is kept at a minimum. According to a talk Peters delivered at the 2010 OASPA conference, institutional memberships now comprise about 1%–2% of Hindawi’s total revenue.
The University of Calgary was one of Hindawi’s first institutional members, joining the program in 2008. Andrew Waller, licensing and negotiation librarian and OA librarian at the university, says, “We felt it was a fit because we had a fair number of faculty members publishing in Hindawi journals, and this just made it easier to administer.” At the same time, the university set up a central OA fund, out of which the institutional membership was paid, to cover publishing costs for its faculty.
Peters acknowledges that the program has its limitations, primarily in the form of fluctuating costs from year to year. Once a library has paid for an institutional membership, researchers have no barriers to submit their work, and the increased submissions drive up the annual membership cost.
However, Waller doesn’t feel this is a major disadvantage. “In the library world, we’re used to dealing with annual subscription prices changing. If I calculate the money I pay to all our OA publishers and divide it by the number of journals we receive, costs [with OA] are in line if not cheaper than if I’d been paying for subscriptions.” Waller also points out that under the Hindawi institutional membership model, an institution with a rising number of submissions will actually pay less in any given year because the flat fee is based on the prior year’s total submissions.
The jury is still out on the long term. Author Richard Poynder, who writes extensively about OA, says via email, “Open access is set to become the dominant way of publishing scholarly papers, but there is no evidence that it will reduce costs and so solve the fundamental affordability problem confronting the research community. Moreover, in a web world, something very different to the journal as traditionally understood is needed, and yet, there is no real evidence that OA publishers are offer ing this, or are likely to offer it in the future (since it is not in their financial interests to do so).”
Continuing to Innovate
Hindawi continues to experiment with ways to better serve its customer, the author. “One important factor for authors is the speed of the peer-review process, and we are taking active steps to try to reduce the delays as much as possible,” says Peters, most recently with the introduction of the International Scholarly Research Network (ISRN.) It’s a series of peer-reviewed, OA journals designed to provide a fast peer-review process for all submitted manuscripts; the goal is to win now the average review time down to 3 or 4 weeks.
Under the ISRN model, which Hindawi first started to use only a few months ago, a topic-specific editorial board of 100–300 members is convened; members must commit to performing the initial review of a submission within 2 weeks of receipt. Each new submission is reviewed separately by five editors who each provide their initial feedback. After reviewing these reports, editors decide whether to accept the manuscript.
The company also continues to move into new fields such as social sciences and is acquiring journals as they come up for sale; three new titles were acquired in the past year.
Peters says the last 6 months have been particularly encouraging for proponents of the OA model. “The OA publishers who are still around are serious; BioMed Central and PLoS are publishing very well-regarded journals. OA isn’t outside of the publishing industry anymore. It’s part of it.”