COMMENT BY: U.S. Department of Agriculture
SUBJECT: Against proposed revision to OMB Circular A-110
DATE: March 9, 1999

To: F. James Charney, Office of Management and Budget

We have reviewed the proposed revision and it raises a number of concerns. Our interpretation of the language is that if a research publication authored by a non-federal employee is used in any way in developing Federal policy or rules and if any part of the research upon which the publication is based was paid for, in whole or part, by a research grant or agreement with the Federal government, then the original research data must be made available to any individual who files a FOJA request by the awarding agency. If that interpretation is correct, then this is a very broad proposal that would be extremely difficult and costly to implement for a number of reasons, outlined below.

The Forest Service is a very decentralized organization. We get several thousand FOIA requests each year. Most are filed with the 160 field offices of the National Forest System. Line officers at those field locations are responsible for developing and implementing policies on the 191 million acres of national forests. The National Forest System line officers are striving to improve their policies and management by making them more science-based. To do that, they have been encouraging their technical specialists (e.g., hydrologists, soil scientists, wildlife biologists) to bring peer-reviewed research results more aggressively to policy and decision making processes. In doing so, the line officers and technical specialists are increasing their reliance on university research and in tramural Forest Service research and development results. In turn, we are placing greater emphasis on the importance of subjecting all our results--whether obtained through in-house research or in collaboration with external research partners at universities and interest groups--to independent peer re view. To us, passing independent peer review is the ultimate test of results' veracity and quality.

The proposed rule would impose a significant additional burden on the Forest Service and our re search collaborators. It would require establishment of procedures to archive data over the long term and creation of a major new data base to track publications based on research funded by the Forest Service and link those publications to the data archives. While it is not yet possible to accurately estimate the costs of implementing this rule, we expect it to cost several million dollars to establish the process and over a million per year from the Research account to maintain it. The annual costs of responding to the increased number of FOIA requests expected by the National Forest System are also likely to be several millions of dollars, which will reduce the funding available for resource management.

We have the following specific comments on aspects of the proposed rule:

o We believe that a clause should be added to Circular A-110 requiring citation of the granting agency and the grant number in any research publication funded, in whole or in part, by a Federal award. Most research publications do not now mention whether or not the work was funded by an Federal award. The burden should rest upon the FOIA requestor to document that the re search publication was funded by a Federal award so that they may make their FOIA request to the appropriate agency. Without requiring authors supported by Federal awards to cite the granting agency and grant number as part of their research publication, citizens lack accurate information to file a FOIA request with the appropriate awarding agency. If the granting agency and grant number are not shown, FOIA requests will probably be filed with the agency developing the policy or rule. This will impose a substantial and unwarranted burden on the agency to deter mine if they, or some other agency, supported the research with an award. Modifying Circular A-110 to require citation of the granting agency and grant number, if any, would simplify the FOIA process for both the citizen and the agency developing the proposed policy or rule.

o We recommend language be inserted in Circular A-110 that requires the FOIA requestor to directly contact the agency that funded the published research results, not the agency using the published results to develop policy or the rule. We believe that a research grant or agreement creates a contractual relationship only between the award recipient and the awarding agency, not all agencies of the Federal government. Our view is that only the agency with the contractual relationship should be held responsible for responding to FOIA requests for data developed as part of the relationship. We reject the idea that any Federal agency should be held responsible for obtaining data from a grant or agreement recipient just because a relationship existed with another Federal agency.

o We applaud the language in the proposed circular revision that limits the data that can be re quested under FOIA to data for research whose results have already been published. We do not believe that unpublished research results and data should be eligible to be requested under FOIA. It is essential that "published" be clearly defined in the circular, so that a clearly understood distinction exists between results whose data must be provided under FOIA and results whose data must not be provided. We recommend three necessary criteria for defining published:

(1) The results were peer-reviewed prior to publication. Our view is that research whose results have not undergone peer review prior to publication lacks sufficient rigor to justify development of policy or rules.

(2) The results document an entire study. Our view is that interim results from research still in progress--even if published--must be regarded as tentative, subject to change as more data are collected. Therefore, we believe that interim results are unsatisfactory for developing policy or rules and the proposed rule should exempt interim results from FOIA requests.

(3) The results appear in a regular periodical or as part of a numbered series. Our view is that results published in the so-called "gray" research literature, such as symposium proceedings, white papers, or other irregularly appearing documents do not meet the quality criteria expected of regular periodicals and numbered series. Also, we believe that published abstracts--even those which are peer-reviewed, document a complete study, and appear in regular periodicals or as part of numbered series--are always too brief to be useful in developing policy or rules. Consequently, we believe that data supporting a published abstract should never be subject to FOIA.

o We believe that a requirement be placed upon developers of policies and rules to keep a complete list of the peer-reviewed, published literature considered during the development process. Currently, many policies and rules are developed without preparing a comprehensive bibliography of the publications considered. Unless and until an accurate bibliography is required as part of the record of decision, no solid basis exists for deciding whether or not a FOIA request filed under this proposed amendment to Circular A-110 is legitimate. This current situation will force agencies into the very difficult position of determining, ex poste, what published literature was or was not used in developing the policy.

o The proposed circular language should be clarified to limit FOIA requests to data for research supporting the policy or rule option ultimate adopted. Often during policy or rule development, a number of options are prepared and considered. A change in the proposed circular language is needed to clarify whether or not published research results used to develop a policy option not ultimately selected, or research considered and then rejected as not applicable to the process, are also part of the pool of published material "used ... in developing policy ...". We believe that limiting FOIA requests to research publications germane to the final policy or rule ultimately adopted is a reasonable limitation. Our concern is that FOIA requests may be overly broad, seeking, for example, "All research which was considered in deliberations on developing policy X." Fulfilling such requests in a timely manner and at reasonable cost would be virtually impossible, in our judgment. Particularly during discovery phases of litigation, we envision receiving requests like this and fear that summary judgements could be issued against the agency for nonresponsiveness should broad requests like this be permissible. We recommend limiting the FOIA requests to the research supporting the option ultimately adopted.

o We believe that substantial investments in archiving data will be required by this circular revision. We are particularly concerned that much of our extra-mural research is conducted by graduate students. Their tenure on campus is finite. Without a data archive, recovering dissertation or thesis data from students after graduation would require tracking them down and depending on their good will to furnish their data. Although university offices devoted to alumni affairs usually have current addresses, obtaining the data in a timely fashion after successfully contacting the graduate is highly problematic. To us, the best way to solve this problem is to require all award recipients to archive a copy of their data before receiving the last award payment; Al though we see merits in having the archive at the university (to better integrate the data with other research underway at the university) there are also merits to having the archive at the funding unit level (to better integrate the data with other research underway by the granting agency). Who should be responsible for the data archive is a matter deserving more discussion with our university partners.

o Because much forestry research is multi-year research whose results remain useful for extended periods of time, the burden of maintaining a suitable data archive will rest more heavily on us than on other fields of science. For example, strip mine revegetation research results published in the 1960's and 1970's are still relevant today for developing policies and rules. Because we successfully solved a number of mine revegetation problems using trees two decades ago, some of the units who conducted that research have been closed or redirected into other reforestation problems and many of the authors have retired. Where those data still exist, they are archived at the Federal Records Center. Recovering them to respond to a FOIA request would be exception ally expensive. Holding universities or our agency responsible for providing data from research completed a long time ago is simply untenable.

In summary, we believe that reliance on the peer review process prior to publication is the best means for determining the quality of research results and creating a warranty of fitness and appropriateness for citizens and staff engaged in development of policy and rules. If the results are published in a peer-reviewed journal, we believe that the publication speaks for itself and the authors should not have to respond to FOIA requests. We believe that the proposed modification of Circular A-110 will create a huge administrative burden on the Forest Service and our cooperating universities and add very little value to policy debates or discussions of proposed rules. Accordingly, should the language from the FY 1999 Omnibus Appropriations Bill directing OMB to modify Circular A-110 be repealed, we heartily recommend that the circular revision process be halted immediately, or if repeal comes after the final circular revision is adopted, that the circular revision be rescinded immediately.

ROBERT LEWIS, JR.
Deputy Chief for Research & Development