Comments of Yale University on
              Implementation of NIH's Public Access Policy




	Thank you for the opportunity to comment on the implementation of NIH's Public Access Policy.

	Yale University agrees with the aims of the NIH's Public Access Policy.  We favor widespread access to the results of federally sponsored research and the intellectual resources of universities.  The established traditions of the scientific community, which demand that research results be published to enable other scientists to test and replicate the findings before they become established fact or working hypotheses, provide broad dissemination of scientific findings.  In addition, Yale has taken specific initiatives to expand access.  Yale has played a leading role with the United Nations in creating programs for very low cost subscriptions to scientific journals through Health InterNetwork Access to Research Initiative (HINARI) and Online Access to Research in the Environment (OARE).  In addition, Yale has launched its Open Yale Courses initiative which is placing entire courses online for use by anyone in the world.  In all of these efforts we have sought to expand access while preserving the strengths of the existing system of scholarly publishing.
	
	Yale has also taken steps to comply with the NIH Public Access Policy.  We have advised all recipients of NIH funding about their obligations under the policy; we have provided model copyright agreements for use by faculty; and will be providing, on an ongoing basis, comprehensive education and support for all university faculty.

	In taking those steps to implement the NIH policy at Yale, we have identified a few issues of concern that we wish to raise with the NIH.

	Alternative Models.  NIH has requested suggestions about alternative models.  The Public Access Policy is awkward in that it imposes a burden on NIH-funded investigators when it may be more efficient for NIH to negotiate directly with publishers about the distribution of articles through PubMed Central.  NIH is, in essence, using the means it has readily available - conditions on grant funding, which investigators cannot avoid - to expand access to scholarly articles.  This approach is expedient but it may not be the best approach.  We would encourage NIH to have constructive conversations with publishers about how to simplify the process for submitting papers to PubMed Central.

	For example, faculty who submit papers to the more than 300 open access journals listed on the NIH website have the option of requesting the publisher to submit the article to PubMed Central.  Faculty can make the request at the time they send a manuscript to the journal for review; no further action is required on their part. The submission process for all other journals is significantly more burdensome on faculty.  Faculty will appreciate the ability to fulfill their obligations in one step, at the time they submit a paper to a journal, and to avoid any further obligation that may come due months later.  We recommend that NIH undertake serious discussions with publishers to achieve a similarly user-friendly option for submissions to any journal.   

	Limitations on Institutional Oversight.  It will be difficult for institutions to ensure that their investigators comply fully with the Public Access Policy.  We recognize that it is common for NIH and other funding agencies to require institutions to ensure that investigators as well as subcontractors comply with various federal rules, such as human subjects rules, that govern the conduct of research.  The Public Access Policy differs from those requirements in at least three aspects, which will make it difficult for institutions to ensure that investigators are in full compliance.

	First, the conduct that is regulated by the policy - reporting of research results through journal articles - often occurs after the research award is completed and in some cases after the author has left the institution.  Some former investigators will be at other institutions; some may be independent; some will be located outside the country,  This would hold true for faculty and especially for postdoctoral fellows and students.  In contrast, most of the policies that govern NIH-funded research are applied and monitored during the course of the research project or are fulfilled soon after a project's conclusion.  In applying those rules an institution can exert considerable leverage over the investigators to compel them to comply.  In the case of publications, an institution may have little or no leverage over authors who have left, will generally not know that the author is submitting a manuscript for publication, and may not even know where the author is.

	Second, the Public Access Policy governs contractual relationships to which the institution is not a party.  Contrast the Policy with the way that funding agencies promote integrity in research, where they do not have direct control over how faculty conduct and present their research.  Rather than imposing a requirement on institutions that governs the scientific communications, the federal policies establish standards for how the institutions manage allegations of research misconduct.  Institutions have control over that process, and can fully comply.  

	Third, the Public Access Policy puts institutions in the middle of investigators' decisions about publication, which are of course close to the core of academic freedom.  For institutions to exercise adequate oversight to ensure that the research publication agreements comply with the Policy, they must monitor or even participate in publication decisions of faculty, fellows, and students to an unprecedented degree.  This is a marked departure from other policies and conditions that funding agencies impose on research.

	By extension, these same issues impact on institutions' ability to satisfactorily monitor their sub-awardees to which these requirements must be passed on.  Institutions, in these cases, have even less leverage or ability to monitor author-publisher relationships.  Because the prime awardee is responsible for the actions of the sub-awardee, institutions may face additional liabilities which seem out of proportion to any harm done. 

	For these reasons we believe it will be difficult for institutions to guarantee full compliance with the Public Access Policy by authors.  Institutions can be held accountable for establishing clear policies about submission of articles to PubMed Central, providing technical assistance to faculty, and requiring faculty to certify that they will comply with NIH's policy.  We recommend that NIH provide a "safe harbor" for institutions which have met those conditions and thus demonstrated that they are making a good faith effort to comply. We believe that such a safe harbor would protect grantees from consequences that are unrelated to their own neglect without absolving them of the responsibility for ensuring, to the maximum extent possible, the success of the new policy.

	Creating a safe harbor may also help to clarify the ambiguity about consequences of failure to submit an article to PubMed Central.  Does NIH intend to demand that federal funds be returned for failure to comply with the Public Access Policy?  That penalty would seem heavy-handed, especially since the compliance "tail" would have no end date - any award could be placed in noncompliance many years after it closes if a former institutional investigator fails to negotiate an adequate publication contract.  We would contend that institutions which qualify for a safe harbor should not be subject to financial penalties.

	Monitoring and Cost of Compliance.  NIH has requested comments about monitoring of the Public Access Policy.  We recommend that any requirements for monitoring and reporting be as efficient and streamlined as possible.   

	We also recommend that NIH, or the National Academy of Sciences, analyze at least two issues as the Public Access Policy is implemented.  First, what is the effect on scholarly publishing?  Have some journals become economically unviable; have investigators migrated to or away from certain journals or forms of scholarly publication?  
	 
	Second, what is the incremental cost for institutions that receive NIH funding?  What is the cost associated with institutional compliance, including training; the oversight of subcontractors; and the acquisition of rights to submit articles to PubMed Central?


	In conclusion, Yale believes universities should make the results of scientific research widely available to the public.  We recommend that NIH clarify important issues about the implementation of the Public Access Policy, especially the extent to which institutions would be held accountable for the actions of individual investigators.  As we note above, the Public Access Policy differs from other conditions imposed on NIH grantees in that compliance is not fully within the control of the grantee.  Providing a safe harbor for institutions that make a good faith effort to comply would help to address those concerns.  We also recommend that NIH continue to explore whether there are more efficient ways to make scholarly articles available, including negotiating directly with publishers.  Finally, we recommend that NIH evaluate the implementation of the Public Access Policy. 

	Thank you for the opportunity to comment on the Public Access Policy.



Public Access Policy
page 3