Briefing on Data Access

February 26, 1999

Wrap Up

GENERAL QUESTIONS TO THE PANEL

Q [Pete Leon, American Association of Engineering Societies]: Ms. Casey, would you address our concerns about the exemptions, specifically medical privacy and intellectual property, and also the cost burden to the agency and the grantee. We would like to be assured that if there is a fee charged, the agency and the grantee get the money, not the Treasury.

A [Kathy Casey]: Cost was an issue that we saw as very legitimate. We expect that ultimately the costs assumed by the recipient will be reimbursed. Regarding medical privacy, our expectation is that there will be some provision in FOIA for that, but with FOIA as the basis, we are not trying to expose what might be considered private or medically sensitive information.

Q [Eric Fischer, Congressional Research Service]: There is a concern I would like to raise having to do with environmental issues. Let us suppose that the OMB revision is drafted as promulgated, and suppose, for example, that a scientist studying a rare and endangered species of plants through a grant from the National Science Foundation publishes a paper on that research. Further suppose that that paper is cited in some document of the Department of the Interior. Would the provision in the law permit a collector of rare plants, or a group of collectors of rare plants, to obtain the location of the scientist's work for its own purposes of collection or resale?

A [Mary Ellen Sheridan]: From the grantee organization's standpoint, if we were asked by NSF (National Science Foundation) to turn over the underlying data, and we currently have no way to insert a barrier to vett that, we would have to comply. So, it is undoubtedly true that the researchers, in their lab notebooks, would identify the sites of the rare plants. From the university's standpoint, we would probably have to turn that over, and then we would look to the agencies to find a way to protect that information that could do public harm through public access and dissemination.

A [Wendy Baldwin]: There is no public harm exception. If one could demonstrate that the release of certain data could cause a commercial harm, then the agency would be in a position to redact that information before making it available. But one would have to make the case, and the agency would need to have the original data to make that determination. Plus, we would have to hold that data, and there is concern that this identifying information or proprietary data would become vulnerable under some other request to a federal agency for data. We certainly have had cases where there has been reach into the agency for data that we would not normally have to provide.

A [Jean Fruci]: There is also a question about what is the definition of "data," what is the nature of the study, and what is being requested. In the case where one plans to collect, maybe not even whole-plant, but leaf and plant samples that one would dry and preserve, there is a question as to whether those actual physical samples themselves might be considered the data that would then have to be given to the agency. In some cases, data may in fact include some physical entity that we do not normally think of, a specialized culture or a plant or soil sample.

A [Kathy Casey]: The question, from my perspective, is whether one is trying to identify some legitimate proprietary interest that is not somehow already addressed, either within the context of FOIA or otherwise.

Q [Eli Schwarz, American Association for Dental Research]: Would the panel explore the possibility of getting a little bit more time to explore the problems around this Circular? I understand there is a 60-day reply mechanism and that there will be a second notice in the Federal Register. How possible is it to have the Institute of Medicine, or a similar body, actually properly study these problems, and is there such a thing as a "time-out" in a situation like this?

A [Marc Garufi]: I believe we are out of "time-outs." We will proceed as proposed with the first and second Federal Register notices, and we will discuss the comments with the policy officials at OMB as well as with the federal agencies. At the same time, it is our obligation to enforce the law. I hope that individuals and organizations take the opportunity to use the comment period to raise questions.

A [Jean Fruci]: Are you saying that you are anticipating a second NPRM (notice of proposed rulemaking) with a second comment period before the final rule?

A [Marc Garufi]: We definitely will have a second notice before the final. Will there be a comment period, Jimmy?

A [Jimmy Charney, a policy analyst at OMB]: Let me clarify. The second notice that we must have as part of the Administrative Practices Act will address the public comments that we received in the public comment period. At this point, we are early in the comment period, and we really cannot say what will be in that second notice. All we can assure you of is that this second notice will be issued and will have a response to the comments. And there will, of course, be a comment period following. The final rule will appear in the Federal Register. In this case, because this is not a proposed rule, but a proposed revision to an OMB Circular, there is another step, in which the OMB Circular becomes the final revision and then the federal agencies codify that into that their regulations.

A [Roger McClellan]: It would be unfortunate if the realm of the debate were restricted to OMB comment or to legislative hearings in terms of Congress. This matter is much too serious to be handled that way. It does deserve extensive debate, and very careful and thoughtful study on the part of the Institute of Medicine or the National Research Council.

Q [Rick Weiden, Vietnam Veterans of America]: Our interest is in obtaining access to databases for subsequent analysis in some kind of repository. I was very disturbed to hear that we do not know how to do that. For example, studies of Vietnam veterans exposed to the toxicological battlefield in Vietnam could be made available to other scientists. Some of that data is based on publications that we believe were incomplete, as an example, the Ranch Hand Study. And it is greater access to that data for valid scientific purposes in which we are interested. I appreciate that this may be more complicated than the mechanism suggested to accomplish it. However, members of Congress have not sought to introduce legislation that fosters debate to come up with something more appropriate. What is disturbing about this to me is we believe that public access to this data by other legitimate scientists is extremely, extremely important on delicate issues. While we wait for these studies, veterans are dying at a much faster rate than those who did not serve in Vietnam, or than would be anticipated based on previous generations of veterans. What do you intend to do to change the context of this debate?

A [Wendy Baldwin]: I am afraid there has been some misunderstanding. I am strongly in favor of data archives. We have mechanisms in place to share data. We have mechanisms in place that ensure that data are available with the appropriate documentation so that it is clear at the outset how data will be made available. Those are the kinds of mechanisms we do need to use. That is a far cry from using FOIA as a mechanism to access data from a grantee. There are some types of data that do not lend themselves to archives. I have concern that having data available via this legislation could in fact create some problems with individual archives. Nevertheless, archives remain a very strong tool for making data available.

A [Jean Fruci]: Let me step back a moment to explain why Mr. Brown introduced HR88 to repeal the provision. We agree that FOIA is the wrong instrument. Therefore, the question arises: What is the right instrument? To answer this, we have to clearly define what our sets of problems are. Then we must examine what is available currently to solve them, and what additional tools are needed. The law is moving on an administrative track. The administration has put out its rule, and is receiving comments, and will move down that track. If one were to engage a mechanism, an excellent suggestion would be to have the Institute of Medicine or the National Research Council convene some groups. It would also be useful for each individual funding agency to work with the respective pools of grantees with whom they normally interact to develop some guidelines and organize some workshops with the science community as well. But if we are on a track with a law that is in place and a rule that must move forward, our ability to examine those other things is very limited. If we move forward with this and we put FOIA in place while we begin to examine these other things, we are afraid that there will be much more legal research generated than scientific research in addressing these issues because of the uncertainty of terms like "publication" and "data," as well as other issues. We would rather see the resources channeled in another direction.

Now, with respect to data on Vietnam veterans, much of the data are actually in the VA records, which are governmental records. Is that true or not true? Are there separate records held by individual researchers? There are clearly some medical privacy issues here that must be addressed, as well. The Institute of Medicine has been studying Agent Orange and tracing the effects of it for a long time, and might, in fact, be the best organization to address the specific issues related your questions.

Q [Mark Frankel, American Association for the Advancement of Science]: Mr. Garufi, the second sentence of the legislation refers to the possibility of collecting some sort of fee to help absorb the costs of this. The OMB proposal also acknowledges this possibility, but is conspicuously silent on how that would be determined and how money collected would be apportioned among the parties involved. Is there anything that precludes the follow-up proposal from having an explicit cost-recovery mechanism in it that might cover both agency and grantee?

A [Marc Garufi]: Well, I am not sure. That is certainly a valid issue. Even though the agency can collect it, this does not mean the money goes back to the agency, even though it is the entity incurring the cost. It is certainly too early to suggest how the second notice will deal with that issue, though it is an important one.

A [Mary Ellen Sheridan]: The administrative costs to universities are already capped, and if this were in any way to be interpreted by either the agencies or OMB as an administrative cost to the universities, we would incur the cost and have the financial liability, but have no mechanism to recover our cost. Having the money going to Treasury perhaps frustrates the federal agencies, but the university community would be capped out. So, the real cost of this is a conundrum to us.

Q [Paul Uhlir, National Research Council]: I think this audience generally appreciates the narrow reading that OMB has given to the legislation in terms of limiting it to published data that are used for regulatory rulemaking. At the same time, even if one were to publish the revisions to A-110, further constraining, and perhaps resolving some of the problems that have been raised in the comment period, there is a concern that the plain language of the legislation states "all data by federal grantees," even if reasonable interpretation and limits on the legislation were provided. That is with any amount of funding. If at some point in the future a politically motivated group would seek to obtain a broader reading of the legislation, I think it is very likely that such a challenge would be upheld in court.

A [Kathy Casey]: That is probably the case anytime a circumstance arises where a particular rule or regulation does not completely marry with the clear language of the statute. That is always a hazard of interpreting Congress' intent.

A [Roger McClellan]: Certainly Circulars are different from regulations in that they have even less force in effects since they are instructions to federal agencies. The likelihood that it would be overruled is probably even greater than if it were a rule. But I think you are probably right with your observation.

Q [Janet Shoemaker, American Society for Microbiology]: Are there any protections for national security concerns under the Freedom of Information Act, especially with regard to research that is increasingly being done to protect against bioterrorism or, biological weapons?

A [Wendy Baldwin]: Yes, there are.

Q: The big problem is, as I see it, in small science. We understand that if we do not make our data available, we will not have a field. The reason data are not shared is that we cannot afford the expensive work of creating the archives and then make them publicly available in the way one would want. The question, for me, is this: If FOIA is not the mechanism for accomplishing data sharing, what is? Where will the resources be found and who or what entity will issue the requirement that enables people to make their data accessible affordably? What is the solution here, and is it really realistic to think it is coming?

A [Jean Fruci]: Well, it certainly is not coming if we do not begin to work on it. That is very clear. And I agree that there is definitely a cost factor involved in not only assembling, but also maintaining these databases and putting them in places where others can access them. There is also a policing activity necessary to ensure that people are indeed contributing to this database and that it is as complete as possible. The funding, I think, can and should come from both the public sector and the federal government. There is no reason why we could not allocate federal funding to the design and maintenance of databases in many fields, where needed. Frankly, there are also some very complicated issues with respect to technology changes that have occurred over the past thirty years. We used to store data on cards that were punched and fed into mainframes. Then we advanced to tapes, then to floppy disks of various sizes, and now we use CD-ROMS. Scientists have stored information for years on all these different types of media, and each one of those has its respective lifetime dictated not only by the technology available to read it, but also by the medium itself and its ability to maintain its integrity. These are significant issues; and they are not simple ones. We must begin to work on them through a couple of different mechanisms, one of which could be the National Research Council convening to look at the specific portions of this issue. Another mechanism could involve a broad engagement of the science community to discover the types of databases necessary. Where should they be located? How can they be made most accessible to others? Who will fund them? What level of funding will be required to maintain them? We must address those both through the science community and through Congress.

A [Wendy Baldwin]: I agree with you. An investigator who envisions placing his data in a data archive can incorporate in his grant application the funding required to make the data appropriate and in a condition to place it in an archive or for him to share it. He may offer in his application that he will make data accessible through his Website, what the conditions will be, and what the effort involved in that will be. This is a legitimate cost on that grant. Some issues arise in doing it that way. What happens at the termination of the grant? This is a dialogue in which the science community must engage; it must talk about what kind of data are appropriate for those data sharing strategies, what kind of data are inappropriate, and would they be shared under unusual circumstances. The anxiety in the scientific community comes from the sense that this is a one-size-fits-all strategy that does not really fit the complexity of the research data to which we are referring.

Q: This question addresses timing. Is there a reason that it could not be extended? Is there a time limit imposed? Is this applicable to State governments, which might be operating with some sort of federal mandate either to gather data or to archive data? If one spends federal money to create the archive, but not gather the data, will one be subject to this?

A [Mary Ellen Sheridan]: If one spent federal money in an archive to enhance those data, which is typically the case, one would invoke federal funding for the production of data.

A [Roger McClellan]: The Circular is effective for colleges, universities, and other nonprofits, not for State and local governments. But, agencies, of course, in their discretion, may extend it if they wish.

A [Wendy Baldwin]: However, if our grantee, operating under A-110, is using data from a State health agency, then he has the reach-through. He may contend that the huge expense of generating these data came from the State level, but if there were a federal grant that used those data for analysis, which led to a publication, which led to a federal policy, then the reach-through would apply.

A [Roger McClellan]: I do not want to minimize the concern over nuances of the rule. We really do need to come to grips with this bigger issue, and it will become a more critical one for the scientific community with our ability to handle our data set. No longer is it necessary to be thinking about going into an archive and examining a thousand laboratory notebooks. We are on the threshold of the ability to have the knowledge of that laboratory in a very small piece of electronic media. So, we need to be looking at this in a proactive way, not in a reactionary way. Yes, problems exist, but let us examine the broader range of opportunities here.

Q: Then the opportunity to extend the OMB process?

A [Marc Garufi]: Well, the process is as such for all Federal Register notices. We believe that the Congressional intent was to try to get the first Federal Register notice out within 90 days. I think we came close, although we were a little bit late. We also believe that the intent is to have the revisions codified by the end of the fiscal year. So, that is the timing.

Q: But is that actually in the language in this case?

A [Kathy Casey]: I think that at least in the timing in terms of when it is finished, that is based on discussions on the floor.

A [Marc Garufi]: Correct. It is not in the language.

A [Mary Ellen Sheridan]: I think we need to remind people that the revision is to Circular A-110, and with the last revision of A-110, the agency implementations of that revision took years.

Milton Goldberg

I have listened intently, and it has been an interesting session. First of all, in conclusion, we all agree that the public has the right to see the results of publicly funded research.

Second, agencies no longer have the right to waive their right to access to research data funded by federal grants, or at least they will not when the Circular is revised. This, in turn, permits the government to secure all data under a grant and because of the law, agencies will make decisions as to what is releasable and what is not. Grantees may be hard pressed now to guarantee confidentiality when it is a federal agency, rather than the grantee, that honors the guarantee.

Third, some believe that the revision actually amends the FOIA. Two important exemptions for the universities are as follows: Exemption four, which protects release of trade secrets and commercial or financial information, and exemption six, which protects release of personnel and medical files. In accordance with those who believe that the outright repeal of the law is likely necessary, some believe that it is inappropriate to apply the FOIA to research grants because FOIA does not recognize that data alone, without careful analysis, without sifting, without review and confirmation, may be misleading or even misused. Often the data is secondary to the analysis. Finally, OMB, acting with the best of intentions, will not be able to prevent legal challenges to its rule. Hence, some believe a legislative alternative is appropriate.

My worry is that the revision will be abused. I think it will be used to challenge gray areas of perfectly good data analysis. I worry, too, that industry may not collaborate with universities, fearful that potential commercial application of a product or process may be released before it is appropriate. I am concerned that the administrative burden resulting from this will be overwhelming. I believe that asking OMB for clarification and definitions probably will not be productive, and that the National Academy and the federal agencies should join in an effort to develop data sharing policies with the major research agencies and that revisions to the Circular should be postponed until that time.