oversight

NSF Controls to Mitigate IPA Conflicts of Interest

Published by the National Science Foundation, Office of Inspector General on 2017-06-08.

Below is a raw (and likely hideous) rendition of the original report. (PDF)

NSF Controls to Mitigate IPA
Conflicts of Interest




June 8, 2017
OIG 17-2-008
AT A GLANCE
NSF Controls to Mitigate IPA Conflicts of Interest
Report No. OIG 17-2-008
June 8, 2017
WHY WE DID THIS AUDIT
NSF draws scientists, engineers, and educators from academia, industry, or other eligible organizations on rotational
assignment to supplement its workforce. These temporary staff members, many of whom NSF appoints under the
Intergovernmental Personnel Act (IPA), can have a heightened risk of conflicts of interest while they are working at
NSF. Given the significant involvement IPAs have in NSF’s award and oversight processes, it is critical that NSF
develop and implement strong controls throughout the agency to identify and mitigate conflicts of interest.
In June 2015, we issued a report documenting problems with NSF’s controls to identity and mitigate conflicts of
interest in the context of one IPA’s tenure at NSF. We conducted this audit to assess the effectiveness of NSF’s
controls for identifying and mitigating conflicts of interest for IPAs agency wide.
WHAT WE FOUND
NSF has implemented internal controls to identify and mitigate IPA conflicts of interest. However, some of the
controls could be strengthened, and additional controls may improve NSF’s ability to identify or mitigate IPA
conflicts of interest. Specifically, NSF’s information system does not restrict conflicted parties from accessing
proposal and award information, and rules on submitting proposals while at NSF are not clear or consistently
enforced. In addition, NSF did not always ensure a substitute negotiator was named when negotiating awards with
former IPAs or fully track completion of exit briefings for departing IPAs. NSF has also not completed agreed-upon
corrective actions to strengthen controls over IPA conflicts of interest.
Given the significant involvement of IPAs in NSF’s award and oversight processes, it is critical that NSF develop and
implement strong controls to identify and mitigate conflicts of interest. If NSF’s decision-making processes are
compromised — or appear to be compromised — by conflicts of interest, NSF risks losing the confidence of the
scientific, engineering, and educational communities; of Congress; and of the general public in the fairness of its
selection process for funding projects.
WHAT WE RECOMMEND
We recommend NSF take corrective actions to strengthen controls over IPA conflicts of interests, including reassess
controls to ensure staff do not have access to awards and proposals for which they are conflicted; ensure that staff
obtain exit interviews; and clarify and enforce its rules on the submission of preliminary proposals by current
employees and IPAs.
AGENCY RESPONSE
NSF generally agreed with our recommendations and proposed corrective actions. Regarding recommendation 1,
NSF responded that it reassessed its controls for access to proposals and determined that the risk of IPAs accessing
their own proposals is so low that no changes to controls are warranted. As part of its response, NSF also provided
feedback in the form of clarifications on a number of sections of the report along with supporting attachments.
FOR FURTHER INFORMATION, CONTACT US AT (703) 292-7100 OR OIG@NSF.GOV.
MEMORANDUM


DATE:                 June 8, 2017

TO:                   Joanne S. Tornow
                      Office Head, Office of Information & Resource Management and
                      Chief Human Capital Officer

                      Lawrence Rudolph
                      General Counsel

FROM:                 Mark Bell
                      Assistant Inspector General
                      Office of Audits

SUBJECT:              Final Report No. 17-2-008, NSF Controls to Mitigate IPA Conflicts of Interest

Attached is the final report on the subject audit. We have included NSF’s response to the draft report as
an appendix.

This report contains four overarching recommendations to strengthen controls over Intergovernmental
Personnel Act assignee conflicts of interests. NSF generally agreed with our recommendations. In
accordance with OMB Circular A-50, Audit Followup, please provide our office with a written
corrective action plan to address the report’s recommendations. In addressing the report’s
recommendations, this corrective action plan should detail specific actions and associated milestone
dates. Please provide the action plan within 60 calendar days of the date of this report.

We appreciate the courtesies and assistance NSF staff provided during the audit. If you have questions,
please contact Elizabeth Goebels, Director, Performance Audits, (703) 292-7100.

cc:            Dianne Campbell        Karen Santoro         Donna Butler
               Dorothy Aronson        Daniel Hofherr        Dr. Maria Zuber
               Christina Sarris       Allison Lerner        Jayne Hornstein
               Marie Maguire          Kelly Stefanko        Louise Nelson
               Elizabeth Goebels
TABLE OF CONTENTS
Background ......................................................................................................................................... 1
Results of Audit................................................................................................................................... 3
  NSF Controls for Addressing Conflicts of Interest ..................................................................... 4
  NSF Has Not Completed Agreed-Upon Corrective Actions to Strengthen Controls over
  IPA Conflicts of Interest ............................................................................................................... 10
Recommendations ........................................................................................................................... 11
OIG Evaluation of Agency Response ............................................................................................. 12
Appendix A: Agency Response ....................................................................................................... 13
Appendix B: Objective, Scope, and Methodology ........................................................................ 28
Appendix C: NSF Controls over IPA Conflicts of Interest ............................................................ 30
Appendix D: OIG Staff Acknowledgments .................................................................................... 31



ABBREVIATIONS
CY                     calendar year
eJacket                electronic award jacket
HRM                    Division of Human Resource Management
ID                     PI identification number
IPA                    Intergovernmental Personnel Act
MIR                    Management Implication Report
OGC                    Office of General Counsel
PARS                   Proposal, PI Meeting, Budget, and Reviewer System
PI                     Principal Investigator
Background
The National Science Foundation is an independent Federal agency created by Congress in 1950 to
promote the progress of science; to advance the national health, prosperity, and welfare; and to secure
the national defense. NSF is vital because it supports basic research and people to create knowledge that
transforms the future.

Appointees under the Intergovernmental Personnel Act (IPAs)

NSF draws scientists, engineers, and educators from academia, industry, or other eligible organizations
on rotational assignment to supplement its workforce. All non-permanent appointments are Federal
employees, except for individuals under the Intergovernmental Personnel Act (IPA), who are paid
through grants and remain employees of their home institutions.
NSF’s reliance on individuals appointed under the IPA — hereafter referred to as IPAs — is significant.
As of December 2016, NSF’s science directorates had 171 IPAs in both executive (28) and non-
executive (143) positions. IPAs headed five of the seven NSF research directorates (one was vacant) and
20 of those directorates’ 29 divisions (two were vacant). Only 52 percent of NSF’s program officers
who run the merit review system and make funding recommendations were permanent employees. 1
Because rotating staff generally are active researchers with a portfolio of NSF awards and are employed
by institutions that receive NSF awards, an IPA’s interests, affiliations, and relationships could create, or
appear to create, bias in his or her work at NSF.

Conflicts of Interest

According to NSF Manual 15 Conflicts of Interest and Standards of Ethical Conduct, “A conflict of
interest is a clash between an official’s concern for the public interest and his or her private interests or
allegiances.” NSF employees, including IPAs, 2 must not be involved in handling any proposal, award,
or other matter in which they, a member of their immediate family, a business partner, or an
organization of which they are or may become a part has a financial interest. For example, a proposal
from an IPA’s home institution coming into his or her division for funding constitutes a conflict of
interest for the IPA, and the IPA must excuse him or herself from handling it.

Internal Control

Internal control is a process effected by an entity’s oversight body, management, and other personnel
that provides reasonable assurance that the objectives of an entity — in this case, compliance with
applicable laws and regulations — will be achieved. According to GAO’s Standards for Internal
Control in the Federal Government, an internal control system is a continuous built-in component of

1
  In addition to IPAs, other non-permanent program officers were temporary employees or appointments under the Visiting
Scientist, Engineer, and Educator program.
2
  Because the conflict of interest rules are the same for NSF employees as IPAs, throughout this report, we refer to both
collectively as “NSF staff.”



1      NSF.GOV/OIG | OIG 17-2-008
operations, effected by people, that provides reasonable assurance that an entity’s objectives will be
achieved. However, no matter how well designed, implemented, or operated, an internal control system
cannot provide absolute assurance that all of an organization’s objectives will be met. Factors outside
the control or influence of management can affect the entity’s ability to achieve all of its objectives.

NSF Process for Addressing Conflicts of Interest

NSF has a process for addressing conflicts of interest when an IPA is selected and is brought on board.
These controls 3 are designed to identify and mitigate conflicts before, during, and after the IPA’s tenure
at NSF.

      •    NSF requires that all IPAs are removed from being the principal or co-principal investigator on
           all active awards and proposals prior to their arrival. In addition, the IPA must identify potential
           conflicts and notify the designated conflict official — who guides staff in resolving conflict of
           interest issues — in his or her directorate, division, office, or the Office of General Counsel
           (OGC) when they arise. The conflicts official examines the matter, decides what special
           arrangements are required, and makes a written record that explains actions taken to remedy the
           conflict. In some cases, a conflict of interest rule automatically disqualifies an IPA from handling
           an NSF proposal or award.

      •    During an IPA’s tenure at NSF, he or she must also obtain permission from the conflicts official
           before participating in any matter in which a reasonable person would question the employee’s
           impartiality in the matter. Conflicts officials determine how to handle a proposal when the
           Program Officer or Division Director who would normally handle the matter is involved with
           the proposal or award or if the official has an affiliation that may create, or appear to create, a
           bias.

      •    Within 90 days from the time an employee — including an IPA — begins work for an agency,
           according to the Code of Federal Regulations, the agency must give the employee the Standards
           of Ethical Conduct for Employees of the Executive Branch and one hour of official duty for the
           employee to review it. Additionally, all program officers and officials and executive level IPAs
           are also required to file an initial and ongoing annual financial disclosure statement.

      •    While at NSF, staff members use NSF’s internal proposal processing/grants management system,
           eJacket, to process NSF proposals and manage post award activities. An award’s eJacket file
           includes internal documents such as the merit reviews of the proposal, the principal investigator
           history, and the review record. Conflicts officials set a warning flag that appears in eJacket for
           each incoming IPA to notify the IPA as well as the IPA’s colleagues when accessing an award or
           proposal with which the IPA was once affiliated as a principal or co-principal investigator.

      •    During their tenure at the Foundation, current NSF employees or IPAs may not submit new
           proposals to NSF. According to NSF Manual 15 and the Code of Federal Regulations, NSF will
3
    See Appendix C on page 30 for more information on the internal controls that we reviewed.




2         NSF.GOV/OIG | OIG 17-2-008
            entertain no proposal in which staff would be a senior investigator or equivalent, unless it is a
            proposal for continuation or extension of support for work that pre-dated service at NSF.

       •    NSF requires all departing IPAs to have a conflict of interest exit interview with OGC to inform
            the staff of post-employment restrictions. OGC generally schedules one session of “Exit
            Interviews for Employees Leaving NSF” per month but allows those unable to attend to meet the
            requirement in other ways, such as scheduling a one-on-one meeting with an OGC ethics
            counselor or independently reviewing written exit materials.

       •    As soon as his or her NSF tenure ceases, a former NSF staff member (including an IPA) may be
            a principal investigator (PI) on an NSF award, according to NSF Rules of Practice. However, the
            former staff member cannot directly deal with NSF officials on any proposal or project for a
            period of 1 year after leaving NSF and must designate a “substitute negotiator” to represent him
            or herself and his or her institution in dealings with NSF during this time.

Some conflicts officials take additional actions that, if required by policy, could provide NSF additional
control to identify and mitigate IPA conflicts of interest. These actions include holding pre-employment
discussions with prospective IPAs to identify their potential conflicts and forewarn them of how
conflicts will restrict their work at NSF prior to their acceptance of appointments; holding a conflict of
interest brainstorming discussion with each new IPA at the start of his or her tenure to help the IPA
identify potential conflicts; and sending a memo annually asking staff to document their conflicts and
compiling them into a list.

Audit Purpose

In June 2015, the Office of Inspector General issued a report 4 documenting problems with NSF’s
controls to identity and mitigate conflicts of interest in the context of one IPA’s tenure at NSF. That
same month, the House Committee on Science, Space, and Technology held a hearing entitled “Is NSF
Properly Managing its Rotating Staff?” As a result of the report, NSF agreed to take actions to
strengthen its controls.

We conducted this audit to assess the effectiveness of NSF’s controls for identifying and mitigating
conflicts of interest for IPAs agency wide. In conducting this audit, we reviewed NSF’s control
activities, which are the actions management has established through policies and procedures to respond
to risks and achieve compliance with applicable conflict of interest laws and regulations.

Results of Audit
NSF has implemented internal controls to identify and mitigate IPA conflicts of interest. However, some
of the controls could be strengthened, and additional controls may improve NSF’s ability to identify or
mitigate conflicts of interest. Specifically, NSF’s information system does not restrict conflicted parties
from accessing proposal and award information, and rules on submitting proposals while at NSF are not
4
    Opportunities to Strengthen Controls over Rotator Conflicts of Interest, June 19, 2015




3          NSF.GOV/OIG | OIG 17-2-008
clear or consistently enforced. In addition, NSF did not always ensure a substitute negotiator was named
when negotiating awards with former IPAs or fully track completion of exit briefings for departing
IPAs. NSF has also not completed agreed-upon corrective actions to strengthen controls over IPA
conflicts of interest.

Given the significant involvement of IPAs in NSF’s award and oversight processes, it is critical that
NSF develop and implement strong controls throughout the agency to identify and mitigate conflicts of
interest. If NSF’s decision-making processes are compromised — or appear to be compromised — by
conflicts of interest, NSF risks losing the confidence of the scientific, engineering, and educational
communities; of Congress; and of the general public in the fairness of its selection process for funding
projects.

NSF Controls for Addressing Conflicts of Interest
We tested the following NSF-wide controls and found that NSF has implemented controls to identify,
mitigate, and address IPA conflicts of interest. However, NSF could strengthen or add to existing
controls to improve its ability to address and mitigate conflicts.

Access to Proposal and Award Information

Because the flag that appears in eJacket when accessing an award or proposal with which an IPA was
once affiliated is only a warning that can be clicked past, all IPAs have access to certain information in
eJacket, such as award, proposal, amendments, and budgets, even if they have a conflict with the award.
According to a Division of Information Systems (DIS) official, if a proposal or award is assigned to a
conflicted Program Officer or within the division of a conflicted Division Director, the flag that appears
in eJacket does not prevent that conflicted Program Officer or Division Director from accessing
sensitive merit review data or providing a funding recommendation for that proposal or award.

NSF’s internal proposal processing/grants management system, eJacket, is used by NSF staff members
to process NSF proposals and manage post award activities. An award’s eJacket file includes internal
documents such as the merit reviews of the proposal, the principal investigator history, and the review
record. Conflicts officials set a warning flag that appears in eJacket for each incoming IPA to notify the
IPA as well as the IPA’s colleagues when accessing an award or proposal with which the IPA was once
affiliated as a principal or co-principal investigator. We tested the existence of flags set for all current
IPAs and found only one that did not have a flag set. However, staff can click through the warning
banner and access certain information, such as awards, proposals, amendments, and budgets.

As shown in Figure 1, the warning banner names the person with a known conflict and states review
data is disabled for everyone but the managing Program Officer, Division Director, Division of Grants
and Agreements personnel, and other authorized personnel and that “it can only be accessed by the
managing Program Officer of the managing Organization.”




4     NSF.GOV/OIG | OIG 17-2-008
Figure 1: eJacket Conflict of Interest Warning Banner




Source: eJacket

As part of our audit, we judgmentally selected three IPAs (one Program Officer and two Division
Directors) and requested that they access within eJacket both an award in which they had been a
principal investigator and an award from their home institution. All three of the IPAs were visibly
uncomfortable clicking through the warning banners, each telling us it was something they normally
would not do because they understood the sensitive nature of the merit review information it contained.
In each case, even though the warning banner appeared, the IPA was able to access the jacket. In one
case, the IPA was a Division Director in the same division that awarded a proposal for which she had
been the Principal Investigator, so she was able to access the merit reviews for her proposal. We
interviewed DIS staff and reviewed NSF’s information systems access controls, which confirmed that
Division Directors have access to all proposals and awards assigned to their division, even those for
which the Division Director has a conflict.

Also, according to a DIS official, eJacket does not prevent an IPA who is a Program Officer or a
Division Director assigned to a proposal for which he or she has a conflict from providing a funding
recommendation for the conflicted proposal. An individual who accidentally or intentionally accesses an
award with which he or she is conflicted can compromise the credibility of the merit review process. An
IPA’s assignment to an award or proposal allows him or her access to sensitive information and the
ability to make a funding recommendation; accordingly, it is imperative that Division Directors and
Program Officers are not assigned to awards and proposals for which they have a conflict.




5      NSF.GOV/OIG | OIG 17-2-008
According to OGC, the flagging module in eJacket has limited utility because it was designed for
Committee of Visitors 5 members, not for program officers’ conflicts, which are more complex. The
warning system cannot always be relied upon because it does not allow for all pertinent conflict
information, and institutions can have multiple and changing codes. Rather than seek a fix to this
technology, OGC emphasized that it is the personal responsibility of each staff member to remember
and adhere to their ethical obligations, including recusing themselves from matters involving institutions
and individuals that create a conflict of interest.

Submission of Proposals while at NSF

We found instances where NSF did not enforce its rules governing an IPA’s ability to submit new
proposals while at NSF. In addition, NSF’s guidance does not address whether IPAs or NSF staff can
submit preproposals while at NSF.

NSF’s Proposal and Award Manual states, “Current NSF employees or IPAs may not submit new
proposals to NSF during their tenure at the Foundation.” Both NSF Manual 15, Conflicts of Interest and
Standards of Ethical Conduct, and 45 Code of Federal Regulations part 680, NSF Rules of Practice,
state that NSF will entertain no proposal in which staff would be a senior investigator or equivalent,
unless it is a proposal for continuation or extension of support for work that pre-dated service at NSF.

We found three IPAs who in 2014 submitted a total of three full proposals for new work to NSF as PI
during their tenures at NSF. NSF awarded two of the three full proposals: one for $329,667 after the
IPA’s tenure at NSF ended, and the other for $783,506 on the day the IPA’s tenure ended.

All three full proposals were coded as a “new project” in the NSF awards system. However, despite the
proposals’ lack of references to prior work, OGC said that it viewed two of these proposals as
continuation of existing work. The third proposal was a full proposal NSF invited as a result of a
successful preproposal. Because IPAs are able to submit proposals and claim them as a continuation of
existing work — which is allowed, versus new work, which is not — NSF’s ability to consistently
enforce its rules regarding IPA submission of proposals while at NSF is weakened. Further, NSF staff
may misunderstand the prohibition against submitting proposals for new work as being dependent on
when the NSF-funded work would start. For example, one IPA who submitted a full proposal had
documented her intention to submit a proposal as PI for work that would start after she completed her
IPA tenure.

We also found two IPAs who in 2014 submitted a total of three preliminary proposals (preproposals) as
PI during their tenures at NSF. As previously noted, NSF invited one IPA to submit a full proposal but
did not solicit the other two preproposals. According to the NSF Grant Proposal Guide, some NSF
program solicitations require submission of a preliminary proposal. NSF guidance does not address
whether IPAs or NSF staff can submit preproposals while at NSF. OGC told us that the NSF General

5
 Committee of Visitors are reviews conducted by external experts at regular intervals of approximately 3 years for programs
and offices that recommend or award grants, cooperative agreements, and/or contracts and whose main focus is the conduct
or support of NSF research and education in science and engineering.




6      NSF.GOV/OIG | OIG 17-2-008
Counsel decided in 2002 that the rules do not prohibit submission of preproposals by current employees
and IPAs; however, NSF did not incorporate this exception into guidance.

Because NSF’s guidance does not address preproposals, all employees and IPAs may not be aware that
they can submit preliminary proposals during their tenure at NSF. Additionally, NSF may not provide
equal access to funding consideration for employees and IPAs if some are allowed to submit proposals
for new projects while at NSF and others wait to do so until after they complete their time at NSF.

Ethics Training

Ensuring IPAs attend an ethics course promptly after their start dates is critical. IPAs typically have not
worked in a Federal agency before and may be unfamiliar with government and NSF ethics regulations.
OGC stated that they discuss situations that are conflicts; provide a copy of the NSF Manual 15,
Conflicts of Interest and Standards of Ethical Conduct; and give attendees an hour of time to review the
manual at each new employee orientation for all staff, including IPAs, on their first day.

In addition, staff at the Program Officer level or above are required to attend at least one Employee
Conflict of Interest training session given by the Office of the General Counsel each calendar year (CY).
We found that all IPAs identified as being required to complete annual ethics training met the
requirement in CY 2015.

In 2012, we recommended that NSF ensure employees complete the agency’s annual conflict of interest
training requirement within 3 months of their start date. 6 In 2013, NSF established that all new Program
Officers meet the requirement by taking session I of its Merit Review Basics course, which includes the
required conflicts of interest training, within 90 days of starting at NSF. The NSF Academy is
responsible for the course, including documenting that program officers have completed the course
through its learning management system, LearnNSF.

Of 113 IPA program officers who have worked at NSF more than 90 days as of the time of our test, 105
IPAs (93 percent) have completed the Merit Review Basics session I course and 8 IPAs (7 percent) have
not completed the course. Of the eight IPAs who did not complete the course, one was assigned to NSF
in November 2014, two in September and November 2015, and the remaining five between August and
October 2016. Of the 105 IPA program officers who have taken Merit Review Basics session I, we
could not determine the timeliness of 29 (28 percent) due to their assignment date being overwritten in
LearnNSF. According to NSF Academy’s Education Services Specialist, the Division of Human
Resource Management (HRM) implemented a solution so staff can track timeliness of course
completion more consistently moving forward.

Although not all, most IPA program officers are completing the course on time. For the 76 IPA program
officers for whom we could determine the timeliness of taking the course, 59 IPAs (78 percent)
completed the course within the required 90 days of starting at NSF, 6 IPAs (8 percent) completed it


6
    Report of Investigation (OIG Case No. 11060042)




7        NSF.GOV/OIG | OIG 17-2-008
from 91 to 120 days of starting at NSF, and 11 (14 percent) completed it from 120 to 240 days of
starting at NSF.

Financial Disclosure

All program officers and officials at or above a comparable level are required to file an initial and
ongoing annual financial disclosure statement. OGC can grant extensions to file of up to 90 days. At the
time of our testing for CY 2016, we found that all but two IPAs — who were erroneously not identified
as needing to file — filed the required financial disclosure statement for CY 2016. Of the 194 IPAs who
filed the statement, 41 (21 percent) received extensions of 2 to 90 days to file. Of the 194 IPAs, 183
filed within the original or extended due date, with 2 of the 11 late disclosure statements filed beyond
the 30-day “grace period” that OGC provides filers. For CY 2015, of 233 IPA financial disclosure
statements filed, 102 (44 percent) received extensions to file of 11 to 92 days. Of the 233 IPA statements
filed, 176 were filed (75 percent) within the original or extended due date, with 6 of the 57 late
disclosure statements filed beyond the 30-day “grace period” that OGC provides filers.

Because IPAs can have a higher risk of conflicts, it is critical for OGC to receive IPAs’ financial
disclosure statements timely so it can identify and prevent potential conflicts.

Conflict of Interest Exit Interviews

NSF requires all departing IPAs to have a conflict of interest exit interview with OGC to inform the staff
of post-employment restrictions. These interviews are important because former NSF staff, including
IPAs, are subject to several Federal statutes, including criminal conflict laws, which stipulate three post-
employment restrictions. 7

OGC indicated that all but two IPAs who left NSF since the start of CY 2015 met the exit interview
requirement. Because OGC does not have a clear record of departing staff members’ compliance with
the exit interview requirement, it searched records such as electronic files, exit briefing rosters, and
meeting calendars to produce this information, which we did not attempt to verify.

To meet the requirement, IPAs attend one of OGC’s group exit interview briefings, generally held once
per month, or if unable to attend, can meet the requirement in other ways, such as through a one-on-one
meeting with an OGC ethics counselor or independent review of written exit materials. OGC records
attendance from group exit briefings in its Financial Disclosure eFiling system (eFile), but only if that
person has not yet met the annual ethics requirement because there is only one field in the system to


7
 Specifically, “for one year after you leave NSF employment, you must not represent private parties (including yourself) in
dealings with any NSF official on any proposal, award, or other matter. For two years after you leave NSF employment, you
must not represent private parties in dealings with any Federal official on any proposal, project, or other matter involving
specific parties if the same matter was actually pending under your official responsibility during your last year at the NSF.”
Additionally, “you must never represent private parties in dealings with any Federal official on any proposal, project, or other
matter involving specific parties if you were personally involved with the same matter as an NSF employee.”




8      NSF.GOV/OIG | OIG 17-2-008
track the annual requirement’s completion. OGC does not always record ways in which NSF staff meet
the exit interview requirement outside of group exit briefings.

Before the IPA departs, OGC documents completion of an exit interview by signing the IPA’s
separation clearance form generated by the IPA’s administrative manager. The form includes a checklist
of steps required prior to leaving NSF, such as completing a conflict of interest exit interview. We
looked for the form for 25 non-executive IPAs who had departed NSF in 2015 and 2016 and found a
form for 19 of those IPAs (76 percent).

HRM specialists responsible for signing the form in various places and filing the form acknowledged
that IPAs often leave early without completing the formal exit process. However, HRM is not always the
last office to sign the form, so departing staff do not always return the form to HRM.

Without fully tracking IPAs’ attendance at OGC exit briefings, NSF cannot be sure that all IPAs receive
the required training — increasing the risk that IPAs, many of whom are new to Federal employment
and unfamiliar with post-employment restrictions, could violate such restrictions. Also, since IPAs can
apply for NSF funding after they leave NSF, it is especially important for them to understand the impact
of their tenure on such proposals.

Substitute Negotiators for Former IPAs

According to NSF Rules of Practice, as soon as his or her NSF tenure ceases, a former NSF staff
member, including an IPA, may be a PI on an NSF award. However, the former staff member cannot
directly deal with NSF officials on any proposal or project for a period of 1 year after leaving NSF and
must designate a “substitute negotiator” to represent him or herself and his or her institution in dealings
with NSF during this cooling off period. A former staff member without a substitute negotiator can
create a bias, or the appearance of bias, in negotiating with NSF on his or her personal awards and
proposals.

Because the appointment of a substitute negotiator is generally documented in a memo and not a field
that could be electronically searched, we reviewed a judgmental sample of 11 awards made to former
IPAs within 12 months of leaving NSF to determine if a substitute negotiator had been named. We
found that NSF made two awards in 2015 to former IPAs who had not named substitute negotiators
within 1 year of leaving NSF as required. After informing NSF of this, NSF immediately provided
documentation naming a substitute negotiator for the one IPA who was still within 1 year of leaving
NSF.

The NSF Rules of Practice does not specify when the designation of a substitute negotiator must occur
and does not state whether a co-principal investigator is expected to fill that role, absent formal
designation. Therefore, NSF should clarify in its guidance when substitute negotiators should be named.




9     NSF.GOV/OIG | OIG 17-2-008
NSF Has Not Completed Agreed-Upon Corrective Actions to Strengthen Controls
over IPA Conflicts of Interest
NSF has not completed corrective actions it agreed to take in response to recommendations in a 2015
Management Implication Report (MIR) from the NSF Office of Investigations. These actions, if
completed, would strengthen NSF’s controls over IPA conflicts.

     •    Develop enforcement tools to enforce the timeframes associated with ethics and financial
          disclosure requirements (MIR recommendation 3) – NSF is required to give an initial agency
          ethics orientation to all staff. Certain staff, including IPAs, are also required to file an initial and
          ongoing annual financial disclosure statement and to attend annual ethics training. The MIR
          recommended using enforcement tools to ensure NSF staff, including IPAs, complete the ethics
          training and financial disclosures in a timely manner.

          In response to this recommendation, the OGC ethics team consulted with NSF conflicts officials
          in January 2016 regarding what enforcement tools might be appropriate and effective. Conflicts
          officials favored blocking access to eJacket for a delinquent filer of a required financial
          disclosure report. We could find no further action taken beyond this meeting involving the
          development of enforcement tools for the timeframes associated with ethics and financial
          disclosure requirements.

     •    Ensure that all staff understand the negative impact of unaddressed conflicts on the
          integrity of the merit review process (MIR recommendation 6) – In response to this
          recommendation, NSF stated that the OGC ethics team would prepare a Staff Memorandum from
          the Office of the Director to all employees and IPAs emphasizing the importance of upholding
          the highest ethical standards. In March 2017, after receiving our discussion draft report, NSF
          issued a staff memorandum stressing the importance for all Federal employees and rotators at
          NSF to uphold the highest ethical standards.

     •    Ensure FastLane and the Proposal PI Meeting Budget & Reviewer System (PARS) do not
          allow the creation of multiple PI IDs (MIR recommendation 10) – All individuals who receive
          NSF funding have a PI identification number (PI ID) that tracks their award history at NSF. An
          institution’s Sponsored Program Officer can create a PI ID through FastLane, NSF’s website for
          researchers and research organizations to conduct award related transactions with NSF, and NSF
          staff can create a PI ID through PARS, an internal proposal management system. At the time of
          the MIR, more than 1,600 NSF staff had the ability to create a PI ID in PARS. In its response to
          the MIR, NSF stated that it would restrict access to the PI maintenance module in PARS to a
          limited number of NSF staff. During the audit, NSF provided a list of 119 administrative
          managers, HRM staff, and DIS staff who can edit and create PI IDs.

          FastLane and PARS appears to still allow the creation of more than one ID for a PI. According to
          NSF, a person only needs a different email address to obtain a new PI ID. However, we found




10       NSF.GOV/OIG | OIG 17-2-008
        186 records (93 matches) involving an identical combination of last name, first name, and email
        address having more than one PI ID.

        We are concerned a PI could be issued more than one PI ID to circumvent restrictions and
        requirements. If the former IPA applies for funding within 1 year of his or her NSF tenure, the PI
        ID should signal a conflict of interest to alert NSF personnel not to communicate directly with
        that former IPA about funding. Another NSF control dependent on one PI ID is the withholding
        of funding when a PI has a delinquent project report. By using another PI ID, NSF could fund a
        PI with a delinquent project report. Finally, NSF may fund a suspended or debarred PI under a
        new PI ID.

        A 2013 OIG audit on project reporting found “[t]here is currently no automated control to
        prevent assigning a second NSF identification number to PIs with an NSF identification number
        who then transfer to another institution.” In response, NSF modified FastLane in August 2014 to
        include a header on the page where organizations register new users stating, “Please do not
        request a new NSF ID for any user who has previously applied to NSF or received funding from
        NSF.”

        We found 321 PI IDs created since NSF put the prompt in place that involved the same first
        name, last name, and middle initial combination. Although it is possible that some of these IDs
        could involve two different people with the same last name, first name, and middle initial, it is
        unlikely that this is the case for all 321 instances given the uniqueness of the names listed.
        Notwithstanding the warning, NSF does not have a way to prevent a PI from obtaining multiple
        PI ID numbers.

        When NSF learns that a PI has more than one ID number, it can perform a merge to combine the
        two histories of the existing IDs and delete the other. However, NSF does not have a policy
        documenting or communicating this ability to NSF staff or Sponsored Project Offices, nor does it
        routinely or proactively search for PIs with more than one ID.

Recommendations
     1. We recommend that NSF reassess its controls to ensure that Division Directors and Program
        Officers do not have access to awards and proposals for which they have a conflict.

     2. We recommend that NSF:

            •   clarify in written guidance its rules on the submission of preliminary proposals by current
                employees and IPAs;
            •   enforce its existing guidance prohibiting employees and IPAs during their tenures at NSF
                from submitting new proposals to NSF; and
            •   enforce the requirement to designate a substitute negotiator for a period of 1 year after
                leaving NSF.




11     NSF.GOV/OIG | OIG 17-2-008
     3. We recommend that NSF assign responsibility for ensuring that staff, especially IPAs, obtain an
        exit interview as required or otherwise acknowledge their responsibility for being familiar with
        post-employment restrictions prior to leaving NSF. We also recommend that NSF develop a
        process for documenting and tracking compliance with the requirement.

     4. We recommend that NSF follow through in its commitment to take corrective actions to
        strengthen controls over IPA conflicts by:

            •   developing tools to enforce compliance with the timeframes associated with ethics and
                financial disclosure requirements; and
            •   making further system enhancements to prevent the creation of a PI ID for an existing,
                identical combination of first name, last name, and email and to document and
                communicate its process for merging multiple IDs for PIs with more than one PI ID.

OIG Evaluation of Agency Response
NSF agreed with recommendations 2, 3, and 4 and proposed corrective actions. Regarding
recommendation 1, NSF responded that it reassessed its system controls for access to proposals and
concluded that the risk of IPAs accessing their own proposals is so low that no system change is
warranted. NSF estimated that the costs to develop a system control would be greater than any potential
benefit. However, system controls are only one part of the internal control structure. Another control is
needed to ensure that Division Directors and Program Officers are not assigned any award or proposal
for which they have a conflict of interest. Such a control would not require costly changes to an
information system.

As part of its response, NSF also provided feedback in the form of clarifications on a number of sections
of the report, along with supporting attachments. We have included NSF's response to this report in its
entirety as Appendix A.




12     NSF.GOV/OIG | OIG 17-2-008
Appendix A: Agency Response




13   NSF.GOV/OIG | OIG 17-2-008
14   NSF.GOV/OIG | OIG 17-2-008
15   NSF.GOV/OIG | OIG 17-2-008
16   NSF.GOV/OIG | OIG 17-2-008
17   NSF.GOV/OIG | OIG 17-2-008
18   NSF.GOV/OIG | OIG 17-2-008
19   NSF.GOV/OIG | OIG 17-2-008
20   NSF.GOV/OIG | OIG 17-2-008
21   NSF.GOV/OIG | OIG 17-2-008
                                  Attachment A




22   NSF.GOV/OIG | OIG 17-2-008
                                  Attachment B




23   NSF.GOV/OIG | OIG 17-2-008
                                  Attachment C




24   NSF.GOV/OIG | OIG 17-2-008
                                  Attachment D




25   NSF.GOV/OIG | OIG 17-2-008
26   NSF.GOV/OIG | OIG 17-2-008
                                  Attachment E




27   NSF.GOV/OIG | OIG 17-2-008
Appendix B: Objective, Scope, and Methodology
This is one of a series of audit, inspections, and special reports prepared as part of our oversight
responsibilities to promote efficiency and effectiveness in agency programs, specifically involving IPAs.
A previous investigation identified problems with NSF’s controls to identity and mitigate conflicts of
interests in the context of one IPA’s tenure at NSF. We conducted this audit to assess the effectiveness
of NSF’s controls for identifying and mitigating conflicts of interest for IPAs agency wide.

We conducted this performance audit between July 2016 and March 2017 in accordance with generally
accepted government auditing standards. Those standards require that we plan and perform the audit to
obtain sufficient, appropriate evidence to provide a reasonable basis for our findings and conclusions
based on our audit objectives. We believe that the evidence obtained provides a reasonable basis for our
findings and conclusions based on our audit objectives.

Our audit reviewed IPA actions recorded in NSF’s awards databases, such as IPAs’ proposals submitted,
awards, and PI identification numbers, with a focus on IPAs at NSF from 2015 through the time of our
audit. We researched Federal laws and NSF guidance. We reviewed prior OIG reports for findings and
recommendations related to our audit. We conducted interviews with both program officers and
executive level staff in NSF’s science directorates serving as conflicts officials and officials in OGC,
HRM, and the DIS in Arlington, Virginia.

To identify any unmanaged IPA conflicts of interest we spoke with NSF’s Designated Agency Ethics
Official, interviewed eight conflicts officials from six of NSF’s seven science directorates, sent a call for
input to NSF’s conflicts officials, and inquired with our Office of Investigations.

To test the effectiveness of NSF’s controls, we reviewed the completeness and timeliness of CY 2015
and CY 2016 IPA financial disclosure filings and attendance at conflict of interest training using OGC-
provided eFile system reports. Because these records are entered, maintained, and reported by OGC, we
were not able to verify their accuracy independently and, therefore, did not rely on them.

We conducted queries of NSF’s award databases of:

     •    IPAs who are not transferred off awards while at NSF
     •    IPAs submitting proposals while at NSF
     •    PIs with multiple PI identification numbers
     •    IPAs who do not have conflict flags set on their PI ID
     •    Deputy Directors who concur on awards from their home organization
     •    Program officers who make a recommendation on a proposal or an award from their home
          institution

For these queries, we looked at the entire universe and did not sample. We did select a judgmental
sample for two tests that support our findings. To test whether departing IPAs had a conflict of interest
exit interview with OGC, our sample of 25 was an estimated 40 percent of the IPAs who left in a given
year. In testing for substitute negotiators, we reviewed proposals awarded to former IPAs within a year



28       NSF.GOV/OIG | OIG 17-2-008
of leaving NSF in CY 2016 or CY 2017, which was 26 percent of the universe of those awarded since
Fiscal Year 2011. We did not project our sample to the population.

We verified the accuracy of the exceptions using corroborating information from other systems and by
sending our exceptions to NSF for confirmation and explanation.

To determine if conflicted party access to proposal and award information is restricted by the conflict
flag, we asked three current IPAs, selected judgmentally, to access awards in eJacket with which they
are conflicted and to click through the warning banner.

We reviewed NSF’s compliance with applicable provisions of pertinent laws and guidance, including:

     •    NSF Manual 15, Conflicts of Interest and Standards of Ethical Conduct
     •    45 Code of Federal Regulations part 680, NSF Rules of Practice
     •    NSF Policy Merit Review Training Requirements for New Program Officers

We identified instances of noncompliance with these laws and regulations as discussed in our audit
findings.

We did not identify any instances of fraud, illegal acts, violations, or abuse.

We held an exit conference with NSF management on March 16, 2017.




29       NSF.GOV/OIG | OIG 17-2-008
Appendix C: NSF Controls over IPA Conflicts of Interest




30   NSF.GOV/OIG | OIG 17-2-008
Appendix D: OIG Staff Acknowledgments
Kelly Stefanko, Audit Manager; Jayne Hornstein, Senior Program Manager; Elizabeth Goebels,
Director, Performance Audits; Marie Maguire, Deputy Assistant Inspector General for Audits; Elizabeth
Argeris, Communications Analyst; and Ruth Gonzalez, Independent Report Referencer, made key
contributions to this report.




31   NSF.GOV/OIG | OIG 17-2-008
32   NSF.GOV/OIG | OIG 17-2-008