Briefing on Data Access

Implications of theLegislation: A View From the Hill

Jean Fruci
House Subcommittee on Technology

I appreciate the comments from Senator Shelby's staff. There may be a difference in perception between Members who sit on authorizing committees versus Members who sit on appropriating committees. In the House, that division is far more distinct than it is in the Senate. The process through which this was developed also involved authorizing committees, holding hearings, and providing an opportunity for the primary community affected by the proposed policy change to comment publicly before Members of the House and Senate. Neither in the case of the House action that took place in the House Appropriations Committee, nor in the actions that took place in this year's appropriations bill, was that opportunity afforded to members of the science community. We think that some of the issues that have been raised could have been clarified and perhaps addressed in a better fashion than they have been if people had been able to hear all sides of the issue.

There was an earlier version of the provision in which OMB (Office of Management and Budget) was asked to study this issue and to report back to Congress with its thoughts on the implications of moving in this direction and making changes to A-110. We were aware of that particular provision and spoke with a few members of the science community about it. Since it was a study, they felt that they would have an opportunity to work with OMB on input and that OMB would reach out to the science community to ask its opinion. I think there was far less concern over a provision that included requirements for a study than there was later over a provision that required that we not study it.

When we first examined this provision, Mr. Brown, with Mr. Porter and a number of other Members, sent a letter to OMB raising the concern that the science community may not have been paying close enough attention to the provision that had been placed in the Omnibus bill, and that they needed to weigh in. They needed to examine the provision carefully through both a scientific and a legal lens to see what changes, if any, in the way the science community does business would result from this change. One of our concerns related to the many medical studies that are done with human volunteers, and allowing the integrity of the relationships that researchers have established with those volunteers through confidentiality agreements to be maintained. The other concern we had was the issue of public and private partnerships, which have proliferated in recent years between universities and the private sector on campus. Would we be able to continue with the intellectual property rules that were being established through provisions like Bayh-Dole? The third concern had to do with looking at the potential increases in administrative costs for universities, nonprofits and hospitals. We have been hearing primarily how this issue affects universities, but hospitals and nonprofits are also covered by the A-110 provision. To some extent, some of the nonprofit research institutes are quite small, and their resources are far more limited than a major research university's would be to address some of the potential administrative costs that might be incurred in making this provision work.

At that point in time, many of our concerns were theoretical because, although we had seen the language in the bill, we certainly had not seen yet what OMB intended to do in response to the requirement that they make the change in A-110. We wanted to ensure that these initial concerns were at least thought about by OMB in the process of pulling this together, and that the OMB reached out to the science community and to the funding agencies to receive ample input from them.

Since OMB issued its Federal Register notice on February 4, we are continuing to consider that, and Mr. Brown will have some formal comments to submit for the record by April 5. We must look at the OMB proposal from two perspectives: One is what will OMB do with the language already in the bill, so that we are not dealing with a theoretical OMB proposal. Second, in light of the language in the statute and the possibility that someone has a different opinion about what OMB decides to implement versus what is in the base legislation, then we must look to the courts as an area for resolution of any differences of opinion which may arise.

I think since the proposal came out, and since we have had a bit more time to examine it, we continue to have concerns in addition to the three areas I previously mentioned. If we look at the very broad base of our concerns, they really involve two major things. One is that FOIA has been a very good law and was something originally designed to provide citizens access to federal agency records. Second, we do have court cases. In fact, a Supreme Court case, which found that research data produced under federal awards are not agency records. There have been numerous other court cases over time which would suggest that all the underlying data produced under these federal awards, are not, in fact, owned by the federal government. The ownership of that intellectual property rests with the researcher and the institution, not with the federal government. If we are going to change those two things, then we need to think carefully about the implications.

One of the areas that we must examine is the overlap with the Bayh-Dole Act, with which many of you are familiar, and which set up rules for how people patent their intellectual property that is produced under federal grants. One question - we certainly have had a number of patents proliferating from universities and researchers that have received federal funding - is if in fact you do not own these data, how can you patent them? What are the potential conflicts with Bayh-Dole, and will we have to re-examine Bayh-Dole in light of essentially broadening the definition of a federal agency to include any university, hospital or nonprofit that has ever accepted a federal dollar?

We are also concerned, to some extent, not about all the benefits that could arise from implementing FOIA and from data access and data sharing. We are concerned that while a cloud hangs over the question of data ownership, how it can be used and when it might become accessible to everyone else, we will create certain disincentives in the science community for putting information out at an early possible date and sharing it.

We are also concerned about very long-term studies. If one is conducting a short-term study that can be turned around in a year or two, then one can publish that paper soon thereafter. But very long-range studies with very large databases that require multiple investigators and multiple years of investment usually result in a stream of publications over time. If, at a very early stage in the process, someone else can step in and utilize that data, as opposed to having to collect his or her own data, will there be a disincentive, if one does not have ownership over these things, to want to perpetuate long-term data sets? We think it may force a short focus on research and research questions.

As we try now to grapple with the definition of data, I think we are also concerned about the possible rise in legal costs. For this, we have looked at a few of the cases that have already occurred at the State level under state FOIA laws. In some instances, where a State-based FOIA law has been used to try to obtain original data from researchers, the definition of data has been very wide. It includes not just the scientific records, but e-mail, phone logs, and extensive financial background information on the researchers doing work. We are concerned that in some instances, vested interests may use this type of provision for the purpose of harassment, rather than for the purpose of gaining data to expand scientific knowledge. We would like to see some sort of precautions put in that would prevent that type of activity, which would have a very chilling effect on research in certain areas. In fact, because OMB confines the language of its proposal to those studies which underlie federal policies or rules, it has almost guaranteed that the studies that would become the focus of a great deal of public attention would be exactly the ones that would be subject to this provision. This would open them up to scrutiny not only by researchers, but also by various vested interests that might have a reason to want to discredit the work simply for the purpose of allaying the federal policy decision.

We are continuing to look at this particular rule. I think it is very clear that OMB has heard from the science community. I congratulate Senator Shelby because for years we have encouraged the science community to pay closer attention to what we do at the federal level, and he certainly got our attention. It is necessary to stay engaged and involved not only with OMB, but with the House and Senate members to fully explain how this provision will move forward and what effect it could have on the research community. Thank you.

Q: [Dee Ann Divis, Geoinfosystems Magazine]: Geographic databases can be the bedrock for policy decision, but they are not necessarily funded by government or the private sector. In fact, there is an effort to allow the use of geographic data in databases specifically prepared by private entities. If data that are privately developed were used as the basis for a federal rule, why would they not be subject to a sunshine law as well?

A [Jean Fruci]: Well, that is a good question. There are a lot of underlying data to federal policies available to the public. I think anyone who has been around the science community knows the phrase, "publish or perish." In fact, if one does not make his work and his results available to his colleagues through peer review journals in a quality fashion, he will not continue to receive federal funds. But we also collect a lot of information upon which we base federal policies and rules from the private sector. A very good example of this is in the pesticide and toxicology arena in the chemical industry. Much of the data that underlies federal policy decisions is not subject to FOIA because it is considered to be confidential business information. One of our objections here is not that people are data sharing or that there is transparency, but that FOIA is the wrong way to go about achieving this. We think there are legitimate reasons for confidential business information, just as we think that there can be legitimate reasons for not picking through each and every data point that someone has used to underlie some other federal policy. But we would agree that if one pursues this, then it does raise the question, "Isn't what's good for the goose, good for the gander"? We use data from many sources, and if we intend to make it all available through FOIA, that is one thing. But we have not done that here. We have concentrated on data that is strictly available through universities, hospitals and nonprofits.

A [Kathy Casey]: Yes, if private data were relied upon for the purpose of a federal rule, then it would in some reasonable manner be accessible in some general way through FOIA. Going back to the issue of confidential business information or privacy issues, we believe that the process facilitates determining what data would be available under FOIA. The general rule is that if it is used to underlie a rule, it should be subject to FOIA. Then the considerations regarding confidential business information, medical privacy, et cetera would be part of the decision-making process by the agency in determining what ultimately would be made available.

Q [Michael Lubell, American Physical Society]: We certainly are not happy with the language that appeared in the Omnibus bill. On the other hand, as a matter of scientific ethics, our society has always supported the notion that a scientist who publishes something must make available all data upon which that publication was based. Why has Congressman Brown not recommended amending the legislation rather than simply repealing it?

A [Jean Fruci]: We think FOIA is the wrong instrument to facilitate data sharing. We should repeal this particular provision, not scrap the issue. I think people do have some legitimate concerns. Certainly Senator Shelby has raised some of them, and others have raised other legitimate data access issues that the House and Senate should explore. We should go in the regular order: introduce some bills, hold some hearings, bring in everyone who has a problem with the current policies, i.e., the funding agencies that distribute the money, the people who make the federal policies, and the people who have provided the data to set those federal policies. Examine what mechanisms may already be available and work well to address some of those problems. What mechanisms do not exist at all and might need to be created either through a legislative process or through the science community? Until we have an opportunity to examine this more carefully, in the wide range of cases that we have, we cannot design a policy simply by saying, "Let's do it through FOIA."

Q [George Leventhall, The Association of American Universities]: Ms. Casey, you said that FOIA provides protection for proprietary information of commercial value, but Dr. Fruci has discussed whether the application of this revision to A-110 affected the Bayh-Dole law. It has been asserted that the protection for intellectual property of commercial or proprietary value only applies to a commercial entity, and does not apply to a nonprofit entity, such as a university where this research would be done. Does the existence of contradictory opinions create the potential of opening up some significant reinterpretations of both FOIA and Bayh-Dole?

A [Kathy Casey]: I think that is a legitimate question that the process that OMB is going through currently can partially address. That process should be sufficient for the purpose of getting comment from the scientific community, from public policy makers, and from the public. We see that as the starting point on this issue.

Q [Paul Uhlir, National Research Council]: There is another legislation that has been introduced in the last two sessions of Congress, in this case through the House Judiciary Committee, to significantly expand the intellectual property protections for databases. Within the scope of that as it is currently written, the universities would be able to enjoy the protections that that legislation provides, as well as the private sector. If this FOIA provision remains as is, would this not really become an incentive for universities to assert intellectual property protections for their databases created under federal grants in order to avoid the FOIA discovery that might be otherwise forced upon them?

A [Jean Fruci]: That is a good question. It is not an angle that we thought of, but it is something we would have to look at seriously. I think there are a number of things here that could create a disincentive for people to share their data and make it publicly available. One of the concerns that I know people have raised is at what point data might become available and what would be the definition of "publication". As you well know, many researchers go to meetings at which abstracts are published and talk about very preliminary things. If, as soon as one discussed the data it would be accessible, one would not even want to talk about it until it was ready to be published. I think there are a number of issues like what has just been raised in copyrighting as well, that we will have to look at.