Letters to the editor

The NIH open access plan

In a September 14 Op-Ed (“A harmful cure”), Martin Frank and Dr. Jeff Glassroth raise a series of objections to the National Institute of Health (NIH) policy for providing free online access to the results of NIH-funded medical research. But their objections are all based on false assumptions and misunderstandings about the NIH policy:

(1) Messrs. Frank and Glassroth write as if the purpose of the program were to bring peer-reviewed medical research to the lay public. This is untrue. The primary purpose is to bring peer-reviewed medical research to professional researchers whose institutions can no longer afford access to the full range of journals, whose average price has been rising four times faster than inflation for two decades. Access for lay readers is important but secondary.

(2) Messrs. Frank and Glassroth complain that the NIH (the world’s largest funder of medical research) “underwrites only about 10 percent of the research published each year.” What bothers them is hard to discern. It seems to be that users will not get the full picture if they rely only on NIH-funded research. True. But thanks to skyrocketing journal prices, it’s also true that users will not get the full picture if they rely only on the holdings of a medical school library. Instead of praising a gigantic first step toward more adequate access, they criticize its incompleteness, as if no literature should be easy to access until all of it is.

(3) They complain that the manuscripts the NIH will put online will be peer-reviewed but not copy-edited. They cite errors that survive peer review but are caught by copy editors. They could just as easily have cited errors that survive both layers of scrutiny. For example, a study by John Ioannidis in the July issue of “PLoS Medicine” found that “most published research findings [hence both peer-reviewed and copy-edited] are false.” If the problem is making fallible literature easier to access, then all published literature should be harder to access, slowing medical research to a crawl. If the problem is reducing the quality of literature available to the public, then it’s illusory. Free online access to peer-reviewed medical research raises the average quality of free online medical claims. If the problem is depriving the public of the marginal improvements provided by copy-editing, then the solution lies in the hands of publishers, like Mr. Frank, who complain about it. The NIH policy allows publishers to replace the authors’ version of the peer-reviewed manuscript with the published, copy-edited version. Mr. Frank is the Executive Director of the American Physiological Society, which publishes 14 journals. He could solve the problem he cites and set an example for other publishers if he would take advantage of this opportunity afforded by the NIH.

(4) Messrs. Frank and Glassroth say that the NIH and proponents of improved access to research “conveniently dismiss the fact that that there are real costs associated with publishing.” Untrue. No serious advocate of open access, and no one at the NIH, has ever said that publishing is without cost. What the writers conveniently dismiss is the fact that the NIH policy is about archiving, not publishing. The policy does not regulate publishing, does not tell NIH-funded scientists where to publish and does not make the NIH into a publisher. The NIH policy only applies to NIH-funded research that has been peer-reviewed and published by independent journals.

(5) Messrs. Frank and Glassroth criticize the call for free public access to publicly-funded research by comparing research to wheat. “The government also subsidizes wheat growers, but they still sell their grain, and no reasonable person asks those who produce bread from that wheat to give their bread away for free.” This is a very bad analogy. Wheat is rivalrous (to use a term from economics), which means that possession or consumption by one person excludes possession or consumption by others. But knowledge is non-rivalrous. It can be shared by everyone without diminishing possession or consumption by anyone. There is a huge difference, therefore, between giving taxpayers free access to publicly-subsidized wheat and giving taxpayers free access to publicly-subsidized knowledge.

(6) Finally, the writers call on the NIH to stop “divert[ing] its scarce resources” to provide public access to the research it funds. They seem to believe that the public access program is very expensive or that the money could be better spent on new research grants than on a program to make all the other research grants more useful. Both assumptions are false. The cost of the NIH program is $2 million to $4 million per year, which comes to 0.01 percent of the NIH’s $28 billion budget. Study after study has shown that free online access increases the impact of research literature, as measured by citations, 50 percent to 250 percent. Free online access makes NIH research more useful, which is good for science, good for taxpayers and good for health care.

PETER SUBER

Open Access Project Director

Public Knowledge

Research Professor of Philosophy, Earlham College

Washington

Story Continues →