One of the creeds of the open access movement is that free access to literature aides the transfer of knowledge from wealthier, better funded nations to researchers in developing nations. There is little to no doubt that increased access to research results has beneficial reverberations in several directions - but like many hypothetical benefits, they only work well if those on the receiving end can efficiently reap those benefits.

Open access works in a number of ways, but the most common are the author-pays model (referred to as gold open access) and the institutional repository model (part of what is referred to as green open access). The author-pays model has several issues associated with it, namely that the author(s) must secure funds to publish their work. This is often untenable no matter what country you work in or what your funding situation may be. For the repository system to work, institutions or individuals must set up and maintain - at their own expense - servers with an online database of their staff's research results. Both remove costs from the reader, though. To summarize how costs are distributed:

  • Traditional model - Publishers recover costs through subscriptions and toll-access fees from readers. Additionally, many authors must pay page charges or color figure fees. Institutions pay for subscriptions that is limited to on-premise use only by staff.
  • Author-pays model - Authors pay publishers to peer review and publish papers in print and/or online. No cost to institutions or readers. Publishers decide a break even point between costs to operate and number of fees/articles they must collect from publishing papers.
  • Repository model - Institutions or individuals (i.e. self-archiving on a personal website) bear the costs of setting up publicly accessible databases of staff's research results. But, these articles are previously peer-reviewed and published following either of the above models. There are no costs to readers. Theoretically, publishers may lose money if readers choose the free article in the repository over the toll-access fee direct from their website should readers not have subscriptions or institutional access.

Clearly this is oversimplified, but I wish to illustrate the challenges facing researchers in countries - or institutions for that matter - which lack a substantial research funding infrastructure. Under the traditional model, researchers with little funding must seek out journals who offer free or low page charges for peer review and publishing - regardless of the importance, novelty or potential level of interest in their work. Lack to publishing funds also prevents them from even considering author-pays models, with the exception of the very few publishers that offer conditional fee waivers.

But publishing is only half of the research battle. The other half is access. All researchers need immediate updates from their colleagues' work to stay on top of their fields and inform their own studies. Independent researchers and those at less fortunate institutions, which are steadily growing in number, cannot access articles behind paywalls. The open access movement holds these individuals as an example of the prime benefits of open access. It has become a moral rallying cry among established researchers, even recently prompting Genomics (an Elsevier journal) editor Dr. Winston Hide to resign citing, "No longer can I work for a system that provides solid profits for the publisher while effectively denying colleagues in developing countries access to research findings."  Yet, little examination is given to how efficiently they can even access free research results.

Some barriers to accessing literature from the internet in developing countries or rural areas include, but are not limited to, 1) computer access, 2) internet access and infrastructure, 3) the associated access fees and 4) the average population-level internet and computer learning curves. There is many indications that access to computers is on the rise all over the world as the internet penetrates further into formerly inaccessible (or, rather, unprofitable) areas. But even in areas with computer access, such as universities and libraries, they are finding it difficult to afford internet access fees. Likewise, developing countries have difficulty in attracting investment from telecommunications companies who are focused on the bottom line to upgrade their networks for affordable, widespread access - especially outside of metro areas.

One way in which open access publishers can play an integral role in improving access to research in developing countries, low-funding institutions, independent researchers and rural areas is to carefully plan their websites around offering a low bandwidth option to navigate and download webpages and files. With broadband and satellite access costs at least 2-3 times that in much of the developed world, researchers and institutions in developing countries face a delicate balance between the cost to access any given website and the importance and value of the information they will receive. While searching a literature database, such as PubMed or Google Scholar, researchers often spend an inordinate amount of time in search mode while the internet meter ticks away. Thus, bandwidth is an important factor in improving access to the developing world.

This is especially relevant in many African and Asian countries where mobile internet access is increasingly penetrating further into hard to access or less profitable markets than broadband cable/DSL. The International Network for the Availability of Scientific Publications reported in 2003 (pdf):

"Bandwidth in developing countries is expensive. In a report for the Partnership for Higher Education in Africa, Mike Jensen calculates that Makerere University paysabout $22,000/month for 1.5Mbps/768Kbps (in/out), Eduardo Mondlane pays $10,000/month for 1Mbps/384Kbps, while the University of Ghana pays $10,000/month for 1Mbps/512Kbps. These figuresindicate that African universities, outside of South Africa, are paying over $55,000/month for 4Mbps inbound and 2Mbps outbound. These figures are about 100 times more expensive than equivalent prices in North America or Europe."

While costs might have come down during the last decade, mobile internet offers a more affordable, solution for a greater number of people. In fact, mobile access fees often cost less in developing countries, up to 60% the cost in developed countries. Yet, many retail mobile providers limit bandwidth with high fees for going over that limit. Downloading a 1 MB webpage to get to a 5 MB pdf file could be 6% of monthly bandwidth limit for 100 MB/month plan. As a researcher you might be downloading several html, pdf and data files each day. Even for a 500 MB/month plan (which is what I have on my android phone in Sweden), you can quickly approach overage fees.

So what is the solution? Repositories and literature search sites need to be designed for speed with less options and more information. Remove or limit sidebars and show the most highly optimized results with the most important or relevant bibliographic information to enable a researcher to efficiently scan the results to find what they are looking for. For publishers, open access publishers in particular, realize that every megabyte counts when access is limited. While you may be free, the internet isn't. Offer a streamlined version of your site in the countries domain (e.g. for in Tanzania) or make a bandwidth-friendly version of your website available as an top-level option. Reduce image file resolution as much possible. Stop the damn pop-up windows for tables and figures (looking at you PLOS...). Downloading an html file should be easier, faster and less bandwidth than downloading a pdf file. Offer easy to use bookmarking services for readers - either via existing social media services or via an efficient built-in system - so that they can easily save and recover articles that are important to them without downloading further documents. Think about the user's experience, reducing the number of clicks to get somewhere and the size of your files. There are no doubt a plethora of solutions to improve website design and function. Some further technical solutions for senior management, librarians and IT staff are outlined in the reported I linked to above.

Why should one care about making access in the developing world more equitable? While your colleagues abroad might be infrastructurally disadvantaged, they certainly are not intellectually disadvantaged. They need historical literature and up to date research results as much as anyone else does to advance the knowledgebase of society as a whole. Research questions and scientific issues in developing countries may be different in scope than those faced by some researchers in developed countries. For instance, a more common parasite or disease in one part of the world might create a demand for research there that could greatly aide researchers who live in areas where that parasite or disease is less common and less researched - and vice versa. Science and research is often focused on problem-solving and is always a global affair. This goes beyond idealistic and moralistic arguments for aiding access to the developing world. Society benefits with increased participation in - and hence, access to - research.

It also a recognition that bandwidth is a resource that needs to be managed and conserved. While some countries are bandwidth rich, others are bandwidth poor. Optimizing your website and publication files should take into account the lowest common denominator among user experience, not necessarily a sleek design and maximizing options. Information and its retrieval should be at the forefront of an open access publisher's thought-process. Otherwise you can ditch the moralistic argument that it improves access to developing nations, when in fact it costs them thousands to just load your frontpage.