Making a pig's ear of an unscientific free-for-all

Arguments for open-access publishing are simplistic and need to be rethought, argues Rene Olivieri

Why can't debates over science publishing be more like... well, science?

Mr Jones had a farm. One day his animals revolted and chased Mr Jones away. Soon the pigs took over and installed an even more oppressive regime. Sixty years after it was first published, George Orwell's Animal Farm still resonates. A movement that aligns itself under the banner of Open Access (OA) has launched a campaign of liberation from the capitalists. Its arguments and rhetoric are often simplistic and inflammatory.

New technology is the precipitating cause of the revolution. It has certainly transformed the world of academic publishing, prompting a debate on how best to disseminate research. The OA movement believes that access to research articles should be free. This is to be achieved in one of two ways: either authors will all place copies of their articles in freely accessible databases, or they will pay for them to be published in freely accessible journals. It is a powerful movement - but its tenets owe more to defunct economics than to fact.

Let's take each in turn. First, one wing of the OA evangelists asserts that placing copies of articles in freely accessible databases will not harm existing journals. This is self-evidently nonsense. We know that libraries do not have enough money to buy everything their users require. If all (or most) of the content of journals were to be freely available and easily retrievable, why wouldn't they save money by cancelling subscriptions? Publishers are already finding that this is what happens.

Second, another OA faction claims that author-pays publishing will cost less. The Wellcome Trust report on OA calls itself an "economic analysis of scientific research publishing" but the economics it employs is more in the Marx and Lenin mould than in the neoclassical tradition recognised by most economists today. The European Commission recently joined the fray, launching its own review of the market.

The most visible author-pays experiments are those of the US-based not-for-profit publisher, the Public Library of Science, which lives off a multimillion-dollar private grant, and those of the largest commercial OA publisher, BioedCentral, which lives off the capital of its founder, Vitek Tracz. The one detailed bit of real analysis done so far, a survey of the leading American Research Libraries, indicates that most top US research institutions would pay more under an author-pays version of OA than they do in subscriptions. And those who are net consumers, rather than producers of research articles - such as industry - would get a free ride.

Moreover, an increasing proportion of the world's published research is coming from non-Western countries. If their fees are waived because they cannot afford to pay to be published, the charge to Western authors will have to rise accordingly.

Third, take the lament: "There is a serials crisis." The notion is that universities are able to provide access to less material, that they are forced to pay increasing prices for journals and are then restricted in how they use them. Some critics maintain that "big deals", where a library buys a list of titles from a publisher, provide poor value for money because the additional content given to the library is used less.

The truth is that the typical academic library in the UK subscribes to nearly twice as many journals as ten years ago; the average price per journal article has declined dramatically since big deals came on the scene. A recent Joint Information Systems Committee study shows that usage of the electronic journal content owned by UK universities is rising by 40 per cent or more year on year, so the price per use is falling even more rapidly. A recent study by a group from the Centre for Information Behaviour and the Evaluation of Research at City University shows that scientists feel they have never had it so good.

So, is everything rosy on Mr Jones's farm? Not at all. There is more that publishers and libraries could do to support scientific and scholarly communications if they stopped posturing and started working together. We could accelerate the elimination of paper journals, which would mean significant savings. We could aim for near-universal access to research among scholars and scientists in higher education. And we could ensure that research is properly interpreted for public consumption.

The new technology changes much, but not quite everything. Back to Orwell's analogy: we need to harness market forces, not overthrow them, if we want to drive further efficiencies and innovations. We must not be held back by moribund ideologies.

Rene Olivieri is chief executive of Blackwell Publishing Ltd.