Conyers and his critic play fast and loose with the facts

Rep. John Conyers (l.), Prof. Lawrence Lessig (r.) (Source: Flickr.com)

An op-ed piece in The Huffington Post by U.S. Rep. John Conyers (D-Detroit) has heated up a recent policy battle between the congressman and prominent legal scholar and activist Larry Lessig over the publication of scientific research. Lessig has responded and both men have cited my earlier assessment of the situation on Michigan Messenger. But neither is playing it completely straight.

Conyers is sponsoring legislation that would end a policy by the National Institutes of Health (NIH) that requires that all research funded by NIH grants be published in Open Access (OA) journals, where the public can access them. Typically, such information is only available to other researchers through subscription-only journals. Lessig is arguing that that the bill is bad policy and that Conyers is being paid off by the publishing industry to get the measure passed. Lessig may be right on the first premise but the evidence does not support the second.

First, let’s look at the substance of the bill, where Conyers seems to be exaggerating — if not entirely inventing — the potential risks of requiring OA publishing of government-funded research. He argues that requiring OA publishing will undermine the process of scientific peer review that protects against the dissemination of bad and even fraudulent research:

While this approach appears to further and enhance access to scientific works, opponents argue that, in reality, it reverses a long-standing and highly successful copyright policy for federally-funded work and sets a precedent that will have significant negative consequences for scientific research.

These opponents argue that scientific journals expend their own, non-federal resources to manage the peer review process, where experts review academic publications. This process is critical because it provides the quality check against incorrect, reckless, and fraudulent science and furthers the overall quality and vigor of modern scientific debate. Journal publishers organize and pay for peer review with the proceeds they receive from the sale of subscriptions to their journals, thereby adding considerable value to the original manuscripts of research scientists.

The claim that journal publishers “pay for peer review with the proceeds they receive from the sale of subscriptions” is not accurate. Reviewers of submitted papers in the sciences are not paid, nor is there any significant cost involved in peer review. The editors forward a copy of the paper to several reviewers, experts in the field, and they offer critiques and make suggestions. These days there’s rarely even postage involved as most of this is done via electronic correspondence.

Second, and far more importantly, most OA journals do have peer review. The Public Library of Science (PLoS), for example, has the same peer review process for submitted articles that subscription-only journals do. And frankly, no scientist with any credibility is going to publish the results of his or her research in a journal that is not peer-reviewed. The very act of doing so is a stain on professional credibility and would make it less likely that he or she would receive another NIH grant in the future.

Third, we have experience in other fields of science that have long had Open Access publishing, like physics. And studies show that after more than a decade of self-archiving in physics, there has been no effect on subscription rates to refereed journals in the field. Indeed, one study notes that a major physics society that publishes subscription-only journals has helped establish an OA journal, which suggests a cooperative relationship between OA journals and subscription journals rather than a competitive one.

Conyers is wrong on the substance of this bill. Open Access publishing as a condition of receiving taxpayer funds is a reasonable and beneficial policy that allows greater access to the results of that research for the public whose tax dollars pay for it. But being wrong is not the same thing as being bought off, and Lessig seems to be straining to make a case that Conyers was paid off to sponsor the legislation where the evidence does not support such an accusation.

His only evidence for this claim is an analysis by MAPLight that shows the contributions made by the publishing industry to members of the House Judiciary Committee in the 2008 election cycle. Lessig says those figures show that “sponsors of this bill — led by Conyers — received twice as much money from the publishing industry as those on the relevant committee who are not sponsors.”

Well, yes, that’s true. But this is a textbook example of how a single isolated statistic can be highly misleading. There are five members of the Judiciary Committee who are sponsors of the bill and they received an average of $5,150 in contributions from the publishing industry. Conyers received $9,000. But there are 34 members of the committee that are not sponsors of the bill, about half of whom also received contributions from the publishing industry. And three of the non-sponsors of the bill received more than Conyers.

The only reason the average for sponsors is twice that of non-sponsors is because about half the members of the committee have received no contributions at all from publishing interests. If you were to compare the five sponsors of the bill with the top five non-sponsors who have received contributions from publishers, the average of the non-sponsors is nearly twice that of the sponsors ($9,700 vs. $5,150). Nearly twice as many non-sponsors received contributions as sponsors. The data simply do not reveal any correlation, much less a direct causal relationship, between sponsorship of this bill and contributions from publishing interests.

When citing my earlier assessment of the situation on Michigan Messenger, Lessig fails to mention that I explicitly criticized his accusation of this being a “money-for-influence scheme” between Conyers and the publishing industry. And Conyers was correct to point that out.

But on the substance of the policy arguments, the facts are squarely on Lessig’s side rather than the congressman’s. This bill is a very bad idea that will only reduce transparency and accountability in the use of taxpayer funds for scientific research at a time when we should be demanding more openness, not less.

Clarification: The wording above regrettably does not make very clear what current law actually requires. The NIH policy does not require researchers to publish articles solely in Open Access journals; it only requires that all articles that result from research funded with public money be deposited in a publicly accessible database like Pubmed Central within a certain period of time after being published elsewhere. The articles can still be published in a subscription-only journal initially. We regret not making that clear in the original article.