In recent weeks we’ve seen a renewed push to introduce open access to science research publications.

The concept is simple: research that is paid for by public funds should be made freely available, not only to other scientists, but also to the people who actually paid for the research: the tax-paying public.

While open access is certainly a noble and worthwhile endeavour, there are problems with the proposed system that few people are aware of and that could actually backfire for the scientific community.

Publish or perish

The importance of peer-reviewed publications to the career of a scientist cannot be overstated. Quite simply, the number of papers a researcher publishes can be a case of make-or-break on a job application or promotion.

Also, the output of such publications is a key quantity considered when governments assess the international standing of departments and universities.

The process of publishing a refereed publication may come as a surprise to some. A newly submitted article is sent by a journal to several other academics to referee, a task they typically undertake for free. Once the article is accepted for publication, it is sold to other academics (through a journal) who wish to access it, usually through substantial library subscriptions from universities.

Perversely, scientists have to pay “page charges” to publish their articles in some journals, and then have to pay to access their own science through library subscriptions.

This situation is, clearly, far from ideal. Indeed, every scientist I have spoken to thinks open access to research publications is a very laudable goal. But it is not the scientists that are the barrier to the public accessing their science – it is the cost of accessing the journals and data.

And while we scientists would like to freely distribute our science, our work only counts if it appears in a refereed publication, sometimes with copyright strings attached that bar us from sharing freely.

Opening the door

In the UK, the government has declared publicly-funded research should be freely available and, to accommodate this, journals have agreed articles they publish will be accessible to all.

If this sounds too good to be true, that’s because it is.

Because the revenue of the journals must be maintained, the financial burden has been moved from those accessing the science to those publishing it. The solution: charging researchers a processing fee of £2,000 to publish their article

Under this “Gold” model, research will be freely available to all.

Covering costs

So where does this extra cost for publishing come from? You might think the cost of library subscriptions would be equivalently reduced, but universities do not buy individual subscriptions. Instead they purchase complex bundles of journals whose costs are difficult to unpick.

It seems that there will be no spare cash from libraries to pay for these publication processing fees.

In fact, in the UK, the government has announced they have put aside £10 million for the transition into open-access publishing, and to address the processing costs.

Again, if this sounds too good to be true, that’s because it is. The money that’s been set aside for the transition to open access (and to cover processing fees) is not new money – it is money that has been raided from the existing science budget.

Essentially this means fewer research projects will be funded, fewer postdoctoral researchers (the coal-face workers of much of modern science) will be hired and, overall, less science will be done.

Back home

So what about here in Australia?

At the moment, we don’t know how the move to open access will proceed, but the rumours are that we will be in a similar situation as the UK, with no currently identifiable pot to pay for the move into open-access publishing.

But how big is the problem? In 2011, the School of Physics at the University of Sydney – where I work – published more than 550 papers on topics from astronomy and astrophysics, through medical imaging, photonic communications, and the actual education of physics.

If we adopt the UK “Gold” model, this is more than A$1 million that needs to be found to continue to publish their work. And this is one school in one university.

With success rates for Australian Research Council Discovery Project proposals being roughly 20%, and with successful projects receiving only about 50% of their requested funding, the prospect of our science budgets being raided to pay for open-access journals is a worrying one.

Researchers in the UK are coming to terms with the upcoming changes, but the thought of continually lining the pockets of the major publishing houses is making some uncomfortable. There has been at least one initiative to set-up an independent journal which will undertake the refereeing and electronic publishing of articles, all for a minimal processing cost.

Unfortunately, new journals lack “impact”, the historical built recorded of attracting the best and most cited research, and until such new publishing initiatives are recognised by governments as being a truly refereed paper, researchers will be compelled, for the good of their careers, to publish in established journals.

Open options

But there’s a twist in this story.

For a number of fields in the physical sciences – such as particle physics and astronomy – researchers have been making their research public for more than two decades on the arXiv preprint server.

It has become the norm for the vast majority of papers in astronomy to be posted to the arXiv upon acceptance in refereed journals, although unrefereed articles from the shadier side of science also appear.

It is clear that such repositories can be very successful. The case of the Hubble Space Telescope Archive is a key example, providing access to data (rather than papers) captured in the entire 20 years Hubble has been staring at the universe. More papers are written on research undertaken with archived Hubble Space Telescope observations than with freshly acquired data.

Many other telescopes make their data public, after a proprietary period, with the goal of maximising the science that can be extracted from initial observations. An astronomer’s career can be made from delving into archives rather than requesting new observations.

But the key to making such repositories of publications and data a success is investment, especially investment in the infrastructure to allow these to be accessed efficiently.

So, in the end, again, everything comes down to cost. And someone, somewhere, will have to pay in the move to open-access science. This cost will undoubtedly increase when the full vision of not only publications, but also all data derived from public funds, is made accessible to all.

But we have to be careful. Raiding existing budgets to pay for this could potentially curtail Australia’s scientific output.

There seems little point having a superb mechanism for everyone to access Australian science if there’s nothing new to access.

Further reading:

• A collection of Conversation articles on open access