The Lancet > Editorial and review > Volume 361, Number 9354, 25 January 2003

Another loss in the privatisation war: PubScience

On Nov 4, 2002, the budgetary plug was pulled on a small inexpensive project that the US Department of Energy (DOE) called PubScience. The project had grown out of the DOE's own work (by the Office of Scientific and Technical Information) to enable access to current research by its own specialists and scientists.

Science journals provided PubScience with bibliographic information, abstracts, and links in structured form, and PubScience integrated that information into a growing searchable database of resources. Publishers provided information for free, and PubScience organised and presented that information; for the past 3 years, the site had been free on the internet. PubScience was by no means a complete database of all scientific literature. The site depended on voluntary participation by publishers. But the project enabled material from small publishers to be presented along with the big journals. It had no axe to grind, was nonproprietary, and was used by tens of thousands of visitors.

The institutional cost of PubScience--a few hundred thousand dollars--paid for a small staff and some office costs. Because the article data were provided in structured form, developing automated ingestion systems was not dear. So why, in a period when access to scientific literature is a vital worldwide need, was PubScience carefully and specifically defunded? Because the lobbyists of the information industry and a few large publishers made the case for "unfair government competition" to the right ears.

I ought to have mixed feelings about this. Although my employer broadly supports open access to scientific research, I am also on the Board of Trustees of BIOSIS, the premier nonprofit indexing and abstracting service in biological sciences. PubScience could be seen as a "competitor" of BIOSIS, since in theory librarians might say "with PubScience free, why should I pay for BIOSIS?" However, it is clear to me that PubScience was not competing with BIOSIS: it was co-existing in a broad field. Although small, the market for smart online services is by no means saturated--it thirsts for diversity. PubScience was a "societal win", using a pre-existing government process to provide a beneficial service to everyone. It was the public sector at its best, helping expand human understanding and promoting a more informed populace. To imagine that PubScience was a "competitor" is nearly laughable. PubScience only provided a different lens for exploring some of the available resources.

Elsevier, the publisher and owner of The Lancet, produces the other primary competitor--the Scirus search engine--which provides bibliographic (and often full-text) searching on all of its 1700 online science journals, along with millions of other bibliographic records and internet pages. Scirus is currently free, and quite useful. The other touted equivalent--Infotrieve-- relied to a large extent on results provided by Pubscience itself. Competition? Elsevier is a privately held corporation with a consistently high annual profit margin. The company has access to the full text of some 40% of the world's leading science journals. The Software & Information Industry Association, of which Elsevier is a member, lobbied against PubScience since it started.

PubScience was not lobbied out of existence because it was pulling market share from the commercial sector. I fear in the end that PubScience was killed to establish a legislative example and to encourage privatisation of public-sector information resources. The US National Institute of Health (NIH), through its National Library of Medicine, produces PubMed, a deep online resource of life-science and medical journals. Rich research capabilities continue to enhance PubMed. The NIH invests large sums annually in PubMed, whose institutional justifications are much like PubScience's were: to provide systems used for NIH's own work to enhance access to high-quality material for the world.

Was killing PubScience a small skirmish in a larger war? Is the ultimate goal to replace PubMed, indeed any publicly funded access tools, with a few privately run and privately profitable equivalents? PubScience was a small underfunded project with a small base of supporters. PubMed has a much larger audience base, and a more immediately valuable result. So PubMed is a more substantive adversary. But that does not mean it is invulnerable.

My hope is to see our society effectively combat its costliest problem: ignorance. To do so, we must strive to expand access to multiple resources, to enable new forms of intellectual exploration, and to ensure that societal choices are undergirded by the best science and research. Extinguishing PubScience by selectively defunding the few hundred thousand dollars it required was a sad small insult to that effort. I hope it is an isolated instance, but I fear it is not.

This commentary expresses Michael Jensen's personal views and not those of his employer.

Michael Jensen
Upper Marlborough, MD 20772, USA