COMMENT BY: Association of American Universities
SUBJECT:Against proposed revision to OMB Circular A-110
DATE:March 23, 1999

Dear Mr. Charney:

This letter is to provide the comments of the Association of American Universities, representing 62 leading North American research universities, on the proposed revision to OMB Circular A-110 published in the Federal Register on February 4, 1999.

In general, we believe that scientific research data should be shared, under appropriate procedures and with appropriate protections. In fact, the current publication process requires that sufficient supporting data be made available so that conclusions can be challenged and/or replicated. This process has served well for many years, and should be carefully considered before it is changed.

We are very concerned that subjecting all data produced through federally funded research to the Freedom of Information Act (FOIA) will create unintended consequences. The National Science Board and the National Academy of Sciences have both called for the repeal of the 1998 law. AAU also hopes that this law can be repealed or amended so that the serious concerns of the academic and scientific community can be addressed. We do not believe it is possible for OMB to allay these concerns given the problems inherent in the underlying statute. We would support having the National Academy of Sciences study the issue of federal data-sharing policy to determine whether there is a problem with existing practices, and how best the objective of reasonable access to data by interested parties can be achieved without threatening the conduct of good science.

We appreciate OMB’s efforts to narrow the application of the law through two mechanisms: making FOIA requests apply only to data underlying published research findings, and only to data underlying research that is used as the basis for federal policy or rules. We suggested similar limitations in a December 4, 1998, letter to OMB Director Jacob Lew from AAU and the Council on Governmental Relations. However, there is reason to be concerned that these restrictions may not withstand a court challenge. Thus, we remain concerned that FOIA requests may come at any point in the progress of research, and may apply to all federally funded research, regardless of whether it is used as the basis for federal policy or rules.

Medical researchers are very concerned that this law may require them to break their promise of confidentiality to patients involved in clinical trials. Even though FOJA provides an exemption for medical confidentiality, it is the agency, not the research institution, that would make the determination as to what data should be masked. Thus, confidential information, such as the names and behavior of patients, for example, would become an agency record, despite prior promises by researchers not to release the information. Epidemiological researchers, who often publish series of articles concerning data collected over years, are concerned that compulsory premature disclosure of data could discourage the participation, or alter the behavior, of study participants, thereby jeopardizing important research into questions that require longitudinal evidence gathering to answer.

If OMB’s proposal to restrict the applicability of FOIA requests only to published research findings were rejected in court, research findings could be released prematurely with inadequate interpretation. This could be misleading and create problems in many areas of public health and safety. In addition, premature disclosure also jeopardizes the scientific process because many professional journals may be reluctant to publish articles that contain previously available and publicized data. This would hamper the dissemination of findings throughout the research community and make raw data available without subjecting them to the rigorous peer review process conducted by these journals. Also, the data in isolation are frequently of little use; it is the analysis and conclusions drawn from the data that are of use, both to the government and to the scientific community. When requested pursuant to FOIA, is a researcher obliged to provide explanatory material to make data understandable?

Release of research data under FOIA may also pose problems relating to ownership of valuable intellectual property. While we understand that FOIA includes an exemption for trade secrets and commercial or financial information, it is not clear whether this exemption is adequate to protect research data that are proprietary or have potential future commercial value, but that are not derived by a commercial entity. For this reason, we are concerned that these requirements may discourage industry from collaborating with researchers at universities and other nonprofit institutions, where proprietary data may be subjected to FOIA requests. Further, these requirements may interfere with the objectives of the Bayh-Dole Act (Public Law 96-517), which was intended to help speed the commercialization of research results by enabling universities to retain patent rights to inventions developed through federal funding. As with many issues raised by this new law, the final answer to the applicability of FOIA to such data will likely be decided in the courts, creating the risk of significant legal costs and years of delay before these questions may be resolved.

Reaching through to data held at the grantee level by a federal agency via application of FOIA fundamentally changes the nature of the federal grant agreement and could effectively result in a convergence of the grant and contract mechanisms. A grant is not a contractual arrangement, in which the government fully compensates the grantee for the product generated and then has rights to its use and any resulting commercial value. Instead, a grant has historically been considered a leveraging device to create incentives for grantees to release to the public products generated with the scientific findings produced through federal funding. This interpretation of a grant was further solidified through the passage of the Bayh-Dole Act.

We are concerned that subjecting research data to FOIA will subject scientists to harassment from interested parties that may be opposed to the research. Repetitive FOIA requests could be filed in an effort to delay or prevent the research. Complying with these requests is likely to be costly and cumbersome. These requirements could have a chilling effect on research in areas deemed controversial, or that evokes the greatest public interest and concern.

It is not clear how much material will qualify as “data,” and therefore how much and what types of information a scientist will need to provide to comply. Ownership or origin of research data is not easily determined. Data come in many forms, such as writings in notebooks, records of interviews, printouts from machines, entries in medical records, and computer-generated drawings. Data can also originate under many different scenarios. There are data brought by the investigator to the project, data provided to the project from another source, data generated under private funding which are then added to data generated under federal funding, and, finally, data generated in such a manner that it would be impossible to determine what the source of funds was that created the record. In sum, raw data do not always come in discrete packages or in such a manner that it is easy to determine their origin. How will grantees be expected to determine what portion of data is federally funded? Is complete access to data required, even those deemed to be privately funded?

Even if OMB’s proposed revision remained in effect and FOIA requests could be made only with respect to data relating to published research findings, the definition of “publication” is very unclear. Does it apply only to publication in a peer-reviewed journal, or does it include release of preliminary findings on the Internet? What if a scientist participates on an Internet message board? What about verbal presentations or poster displays at society meetings or other gatherings? What if a study is continuing but interim results are published?

We are also concerned that these requirements will discourage collaborations with international research agencies, since other countries will not have similar data-sharing policies in place.

We are very concerned about the ability of research institutions to recoup the costs associated with compliance with these requirements. While OMB’s proposed revision authorizes an agency to charge “a reasonable fee equaling the full incremental cost of obtaining the data,” it does not apparently provide for the research institution to recoup its costs. Further, we understand that present federal policy calls for all agency fees to be turned over to the Treasury, so these requirements may also not enable agencies to recoup their costs.

In addition to these general concerns, OMB’s proposed revision raises the following specific questions:

• What data relate to publication—only those data underlying the article submitted to a journal, or all background work by the scientist?

• Are data defined as those published anytime in the past, or just newly published?

• Will agencies be allowed to determine their own policies as to what constitutes data, and what data must be released, or will there be a uniform governmentwide policy?

• How will it be determined whether data have been “used.. . in developing policy or rules”? Will agencies be required explicitly to identify published materials underlying a rule?

• What if the researcher changes institutions? Which institution bears responsibility for complying with the FOIA request?

• How long must the data be made accessible? Are the data in question subject to release as long as the research project continues to receive federal funding (which could be over a long time period), or, worse, as long as the regulation or policy is in effect?

We have found that the more we review the new law, the more questions arise. It is possible that we will submit additional comments. In addition, as we requested in our December 4 letter, we hope that the sixty-day public comment period can be extended. This could be accomplished by reproposing the revisions to the circular for a second round of comment after the initial comment period, and after receipt and consideration of the first set of public comments.

Thank you for your attention to our concerns.

Cordially,

Nils Hasselmo
President