NATIONAL SCIENCE FOUNDATION OFFICE OF INSPECTOR GENERAL OFFICE OF INVESTIGATIONS CLOSEOUT MEMORANDUM Case Number: A09100080 Page 1 of 1 We received an allegation that a Director of an NSF-funded organization, which receives and considers proposals for research using the organization's resources, had improperly handled a proposal on which he had a conflict of interests. Our investigation found evidence that: (i) the PI of the proposal had information not available to the public; (ii) the Director overrode the recommendation of the organization's review committee and funded the proposal; and (iii) the Director was a current collaborator of the PI. We recommended the NSF Division determine if the affected scientific community would benefit from the Division working with the organization to evaluate and improve its policies and determine if it is in the community's best interests for the project to continue to be funded. The Division agreed to do so. Accordingly, this case is closed with no further action taken. Our report, the Division's decision, and this Closeout Memorandum constitute the documents for the case closeout. NSF OIG Form 2 (11 /02) Date: 19 April 2012 To: OIG From: Subject: Response to OIG investigation A09100080 on the improper handling of a - t i m e award b y - -supports the widely-applied principle that the-directors, as delegated by the awardee managing organizations, have considerable discretion in the allocation o f - t i m e (including explicit authorization to allocate "directors discretionary time") analogous to the authority for DO concurrence of awards at NSF. However,.also considers it essential that the time allocation processes be conducted under appropriate, unbiased policies such that NSF and the community maintain the integrity and trust in t h e - a n d awardee. When authorized to do so,.will initiate consultation with the awardee - t o develop and implement policies and actions fo~along the following lines: (1) Expand-current COl policy to include existing and recent collaborators and extend application of the policy t o - s t a f f including the Director. Staff should recuse themselves if a COl exists .• approval of time allocations will be required if the Director has a COl. (2) Establish a policy and practices that provides all time allocation applicants with access to the same information (such as available technical and programmatic resources and review criteria) . Time allocation committees will be instructed to disregard contrary information contained in any individual application. Any deviations from this policy, for example to develop new observing modes or techniques or new instrumentation, must be documented and require the explicit approval of t h e - Director or-per the COl policy. (3) Establish a threshold for directors discretionary time allocations. Exceptions to this threshold will require the explicit approval o - (4) Any editing or redaction of comments from referee reports prior to conveying them to time allocation applicants must be documented and explicitly approved by management following the COl rules. (5) Implementation of a reconsideration policy following the new COl rules. (6)~0 conduct a review, independent o f - of the scientific merit and handling of the proposal at issue with a view to continuing the allocation of observing time . National Science Foundation Office of Inspector General Confidential Report of Investigation Case Number A-09100080 10 May 2012 This Confidential Report of Investigation is provided to you FOR OFFICIAL USE ONLY. It contains protected personal information, the unauthorized disclosure of which may result in personal criminal liability under the Privacy Act, 5 U.S.C. § 552a. This report may be further disclosed within NSF only to individuals who must have knowledge of its contents to facilitate NSF's assessment and resolution of this matter. This report may be disclosed outside NSF only under the Freedom of Information and Privacy Acts, 5 U.S.C. §§ 552 & 552a. Please take appropriate precautions handling this confidential report of investigation. NSF OIG Form 22b (11/06) CONFIDENTIAL CONFIDENTIAL Background We received allegations that the Directorl of an Organization2 improperly handled a proposal3 submitted to the Organization, and that the Director had a conflict of interests with that same proposaL The proposal was submitted to the Organization on 1 Jun 2009; it was merit reviewed and subsequently discussed by the Organization's review committee (RC). 4 The RC recommended against the Organization allocating any equipment time to the proposal because of several concerns. Instead, it suggested the proposal be resubmitted the next trimester.5 However, the Director provided a disposition to the PI stating the Organization would support the proposaL 6 The Director also provided an edited version of the RC's comments (removing some of its criticisms and its recommendation for resubmission) and copies of the individual reviews, ofwhich he edited one to remove negative comments about designation of the project as a "Key Project" and its request of a large amount of [instrument] time.7 The alleged concerns raise issues related to several Organizational processes including: • Security of Privileged Information-(a) the PI requested "Key Project" designation; and (b) the PI proposed making use of instruments that did not exist but for which a different, confidential proposal was submitted to NSF for funding. Both of these requests could only be made by someone with access to the Organization's privileged information. That they were made by a PI, who should not have such access, raises questions about the protection of privileged information within the Organization. • Conflict of Interests Policy- The Director overrode the RC to fund this proposal using his discretionary authority. His actions were unusual because of the large amount of time requested and the technical concerns of the RC. That he could take such actions raises a question as to whether the Organization's COl policy adequately addresses the professional and personal relationships between Organization officials and applicants. 1 [redacted] 2 [redacted] 3 [redacted] Tab 2, Attachment 1. 4 [redacted] 5 Tab 2, Attachment 3. 6 Tab 2, Attachment 4. 7 5280 hours of the [redacted] over 5 years. 1 CONFIDENTIAL CONFIDENTIAL Summary of Investigation We requested all information related to this proposal from the Organization. 8 The President of the parent organization,9 which is the NSF grantee, provided the proposal, the referee's reports, the RC report, and other relevant informationJO Security of Privileged Information: The PI used the designation of "Key Project" in his proposal: "(h]ere we propose a Key Project for the [Organization]." 11 The proposal was submitted on 1 June 2009. The "Key Science Project" designation was first announced in the Organization's September 2009 newsletter, 12 and was not available for projects until the 1 October 2009 deadline for proposals. In fact, one of the referees specifically commented on this issue: It is to be noted that the authors place these observations in the framework of a key project. While I understand the [Organization] may eventually move in this direction, this has not yet been announced. This proposal hence sounds a bit premature to me. Preemption of a large amount of [instrument] time before the official announcement of such key programs should be avoided in order to be fair to all potential observers.l3 Additionally, the RC summary for the proposal notes: Finally, this proposal is premature as the announcement regarding key projects was not made available to the community for the June 2009 proposal deadline. The designation of a key project shall be made by the RC and not by the proposer(s).l4 The PI's use of this designation clearly indicates the PI had inside knowledge of this designation before the general [redacted] community. The description of "Key Projects" agrees with the RC's comment that the "Key Project" designation "is based on its scientific ranking from the proposal review process, nomination by the organization [RC], and final selection and approval by the organization Director's Review." 15 From the Director's disposition, it appears this project was granted the "Key Project" designation, "... to work through the practicalities required to carry 8 Tab 1. 9 [redacted] 10 Tab 2. 11 [redacted]. According to the Organization, Key Projects are projects that are expected to "have high science impact [redacted] ." These projects have a priority on receiving [redacted] time , as well as a longer amount of [instrument] time. Organization staff is expected "to provide maximum possible support" and students working on Key Projects are encouraged to participate in specialized training. 12 [redacted] ; Tab 1, Attachment 7. 13 Tab 2, Attachment 2. 14 Tab 2, Attachment 3. 15 Tab 2, Attachment 8. 2 CONFIDENTIAL CONFIDENTIAL out this very significant Key Project" 16 despite the RC's reservations and in violation of the stated policy that the RC should nominate a project for this designation. Thus, the Director appeared to bypass the organization policy by designating the proposal as a "Key Project". These actions resulted in this proposal receiving unusually favorable treatment. Another concern, which relates to the allegation that the PI had access to privileged information, is the PI's proposed use of an instrument that did not exist. The PI proposed making use of a special instrument,l7 which the Organization had not announced to its community--even the RC did not know about the instrument. One reviewer wrote "part of the proposal requests time with a new [instrument] system on the [instrument] which has not been funded, let alone deployed or commissioned." 18 The RC also addressed this concern: "This is a proposal to use capabilities which do not yet exist for the [Organization] and are dependent on the outcome of the MRI proposal submitted separate to the NSF."J9 Additionally, several of the reviewers commented that the PI's proposal, without the proposed (non-existent) instrument, was just better than average, and the RC concluded "the backup plan [if the new instrument doesn't get funded, built, and operational in time] is thus much weaker than the proposal as a whole." 20 We concluded the PI's plan to use a nonexistent instrument also supports the allegation that the PI had access to non-public information.21 Conflict of Interests Policy: The Organization's COl policy states: Referees should not evaluate proposals by teams which include themselves, their spouses, significant others, or other immediate family members. Beyond that, we have no formal rules with respect to competitors, past students or advisors, collaborators on other projects, and so on. 22 The Director is and has been a collaborator with the PI prior to, and during, the PI's submission and the Director's dispensation of this proposaL They currently collaborate on a notable research project together, including publishing together. 23 By NSF standards, the Director would be recused from any participation in the evaluation of the PI's proposal because of his professional relationship with the PI. It appears that under the Organization's current policy such relationships do 16 Tab 2, Attachment 4. 17 [redacted] 18 Tab 2, Attachment 2. 19 Tab 2, Attachment 3. 2o Tab 2, Attachment 3. 2 1 In the reviews the Director provided to the PI, discussion about the PI's use of the Key Project designation were removed in one individual review and with two edits to the RC summary. 22 Tab 2, Attachment 6, p. 5. 23 The Director and PI are current collaborators on [redacted] . 3 CONFIDENTIAL CONFIDENTIAL not constitute a conflict of interests because the Director and organizational officials are not collaborating with the PI on the project described in the proposal. Conclusions The overturning of the RC's recommendation, by an Organizational official involving a current long-term collaborator who apparently had access to privileged information, on a proposal with such a large impact on the available community time on the instrument, creates (or reinforces if it already exists) the appearance that award was not objective. These actions run counter to the expectation that NSF, and the Centers and Facilities it supports, provides support to the whole of US science in a fair and unbiased manner. The effect of the Director's decision was significant. Specifically, the RC was concerned about the large amount of instrument time the project requested. It requested approximately 5280 hours, which represents approximately 30% of the total time available on that instrument. 24 Five of the eight referees commented on the requested time. 2 5 One said the total time looked "reasonable" and another noted the time resources were "beyond any request I have seen," while a third described the proposal as asking for "a large amount of time." The fourth reviewer who commented on time noted if results turn out as stated, the time "is a small price to pay," but since that was unlikely, he recommended only 40% or 2000 hrs. The RC as a whole noted, "A proposal for such a large amount of time should have no reservations about the observational plans or the technical aspects" and this one did. It said the [fifth] referee's question "'Is the foreseen improvement ... worth 5000 hours of [ ] time' is left unanswered." 26 Thus, the decision to override the RC for this particular project has a significant impact on the community due to the amount of time the PI will be allocated (approximately 30% of the total time). Decisions such as this should be made as objectively as possible. OIG Recommendation As we have noted, the Organization operates without the standards NSF has in place to protect the integrity of its merit review. We recommend NSF: • Determine if it is in NSF's and the community's best interests for the Organization to continue to operate under its current review policies. Specifically, NSF and the Organization should consider the need for: 24 This estimate is based on past [redacted] schedules where 1200 hours of [instrument] time was a typical trimester load . Thus, 5280 hours over 5 years is 1056 hours per year or about 30% of the total time. 25 The following quotations are from the various reviews; Tab 1, Appendix 2. 26 Tab 2, Appendix 3. We note the Director edited out this comment in his disposition-Tab 2, Appendix 4. As previously noted , the Director also edited out the RC's declination: "Based on the above , we cannot allocate time to this proposal." 4 CONFIDENTIAL CONFIDENTIAL o a more substantive conflict-of-interests policy to ensure an unbiased process for reviewing proposals; o a policy about providing unedited 27 reviews and RC summary (including recommendation) to the PI; o a policy clarifying the Director's role m the revwwmg and funding process; and o a reconsideration policy for the review of questionable funding decisions. o a recusal process that could be relied on to ensure objective evaluation of proposals. • Review the handling of this proposal and determine if it is in NSF's and the community's best interests for the Organization to continue to fund this project and whether any individuals should be recused from further involvement in this project. 27 We are describing the Director's actions as editing, not redacting, the reviews. We are not addressing the simple redaction or strikethrough of material deemed to be ad hominem or not relevant to the reviews. 5
Published by the National Science Foundation, Office of Inspector General on 2012-05-22.
Below is a raw (and likely hideous) rendition of the original report. (PDF)