Title : Number 13--NSF OIG Semiannual Report to the Congress Type : Report NSF Org: OIG Date : January 23, 1996 File : oig13 National Science Foundation Semiannual Report to the Congress Number 13 April 1, 1995 - September 30, 1995 Office of Inspector General Letter to the National Science Board and the Congress This report describes our activities and accomplishments for the second half of FY 1995. Section 5 of the Inspector General Act of 1978, as amended, requires that the National Science Board transmit this report to the Congress within 30 days of its receipt along with any comments the Board may wish to make. We implemented a strategic plan for audits that employs risk assessment principles to maximize the effectiveness of existing audit resources. We recommended several actions that, if implemented, could save the government more than $7 million. We also began reviewing key operational areas that will be used in auditing agency-wide financial statements for FY 1996, when the Chief Financial Officer and Government Management Reform Acts require that we begin auditing financial statements that encompass all of NSF's operations. In this report, as well as in presentations before members of the research community, we continue outreach activities to explain policy and procedures regarding misconduct in science and assist the community in resolving allegation as they arise. We referred six cases involving allegations of misconduct in science to the Deputy Director for adjudication. Linda G. Sundro Inspector General October 31, 1995 Executive Summary Financial Audits Our strategic plan for audits will help conserve audit funds. By using risk assessments principles, we focus on financial irregularities that involve large dollar recovery. The government could save $4.5 million by not paying fees to a nonprofit organization to cover costs that are unallowable under federal cost principles. In our evaluation of the operations of a research vessel in Antarctica, we found three areas where the program could save between $1.2 and $2.6 million. NSF could save over $500,000 by changing the methods it uses to provide administrative services for peer reviewers and other visitors. Fifty organizations used $600,000 in funds designated for participants at NSF-sponsored meetings for other purposes. We are conducting additional audits. We are reviewing key operational areas that will be used in developing agency-wide financial statements for FY 1996. Inspections and Program Oversight We conducted inspections at a state university and a liberal arts college. We also found that an IPA serving as a program officer violated agency regulations by handling a proposal from his home institution. Investigations A senior official at an educational association pled guilty to embezzling $19,598 in grant funds. A U.S. District Court sentenced one small business to 5 years' probation. The Department of Justice is reviewing the actions of other business that received awards under the Small Business Innovation Research program. Misconduct in Science Successful adjudication of misconduct in science cases involves both the determination that deviation from accepted practices is serious and the determination of the degree of serious deviation from accepted practices that has taken place. The concept of serious deviation from accepted practices, as defined by the relevant scientific community's standard of professional conduct, is no less definite than standards of conduct upheld by the Supreme Court for other professions. We recommended that the Deputy Director finds misconduct in six new cases. These cases involve plagiarism, breach of the confidentiality of peer review, failure to follow regulations on the use of human subjects, and misrepresentation of credentials. Table of Contents AUDITS INVESTIGATIONS OVERSIGHT LEGAL AGENCY REFUSAL TO PROVIDE INFORMATION OR ASSISTANCE SIGNIFICANT MANAGEMENT DECISIONS THAT WERE REVISED INSPECTOR GENERAL'S DISAGREEMENT WITH SIGNIFICANT MANAGEMENT DECISIONS LIST OF REPORTS STATISTICAL INFORMATION REQUIRED BY THE INSPECTOR GENERAL ACT OF 1978, AS AMENDED INSPECTOR GENERAL REPORTS EXTERNAL AUDIT REPORTS WITH OUTSTANDING MANAGEMENT DECISIONS ADDITIONAL PERFORMANCE MEASURE Acronyms ASA Antarctic Support Associates ASC Division of Advanced Scientific Computing CAS Cost Accounting Standards CFO Chief Financial Officer DAEO Designated Agency Ethics Official DoJ Department of Justice EHR Directorate for Education and Human Resources GMRA Government Management Reform Act IPA Scientist Employed Under the Intergovernmental Personnel Act IRB Institutional Review Board for the Protection of Human Subjects NASA National Aeronautics and Space Administration NSB National Science Board OGC Office of General Counsel OGE Office of Government Ethics OMB Office of Management and Budget OPP Office of Polar Programs PI Principal Investigator RUI Research in Undergraduate Institutions SBIR Small Business Innovation Research Program SSI Statewide Systemic Initiative Reporting Requirements The table cross-references the reporting requirements prescribed by the Inspector General Act of 1978, as amended, to the specific pages in the report where they are addressed. Requirements Section 4 (a)(2) Review of Legislation and Regulation Section 5 (a)(1) Significant Problems, Abuses, and Deficiencies Section 5 (a)(2) Recommendations With Respect to Significant Problems, Abuses, and Deficiencies Section 5 (a)(3) Prior Significant Recommendations on Which Corrective Action Has Not Been Completed Section 5 (a)(4) Matters Referred to Prosecutive Authorities Section 5 (a)(5) Summary of Instances Where Information Was Referred Section 5 (a)(6) List of Audit Reports Section 5 (a)(7) Summary of Each Particularly Significant Report Section 5 (a)(8) Statistical Table Showing Number of Reports and Dollar Value of Questioned Costs Section 5 (a)(9) Statistical Table Showing Number of Reports and Dollar Value of Recommendations That Funds Be Put to Better Use Section 5 (a)(10) Summary or Each Audit Issued Before This Reporting Period for Which No Management Decision Was Made by the End of the Reporting Period Section 5 (a)(11) Significant Revised Management Decisions Section 5 (a)(12) Significant Management Decisions With Which the Inspector General Disagrees AUDIT The Office of Audit is responsible for auditing grants, contracts, and cooperative agreements funded by NSF's programs. It also reviews agency operations and ensures that financial, administrative, and program aspects of agency operations are reviewed. The Office evaluates internal controls, reviews data processing systems, and follows up on the implementation of recommendations included in audit reports. In addition, the Office assists in the financial, internal control, and compliance portions of OIG inspections. All audit reports are referred to NSF management for action or information. The Office of Audit advises and assists NSF in resolving audit recommendations. The Office also acts as a liaison between NSF and audit groups from the private sector and other federal agencies by arranging for special reviews, obtaining information, and providing technical advice. The Office of Audit provides speakers and staff assistance at seminars and courses sponsored by NSF and other federal agencies and at related professional and scientific meetings. RISK ASSESSMENT AND SURVEYS We continue to devote most of our audit resources to reviewing grants, contracts, or cooperative agreements. This approach focuses our audit resources on external organizations that receive most NSF funds. NSF has overall responsibility for ensuring that certain organizations that receive NSF funds have financial systems that are adequate to properly administer federal awards. NSF is designated as the "cognizant" federal audit agency for these organizations based on Office of Management and Budget (OMB) Circulars A-88 and A-133. We are expected to audit these organizations on a regular, cyclical basis. NSF is the cognizant federal agency for approximately 650 institutions, including not-for-profit organizations (such as education associations and museums) businesses that receive NSF awards, independent centers and facilities, and contractors who perform work for NSF. NSF is not the cognizant federal audit agency for most of the colleges and universities it funds. When the OIG was created in 1989, its predecessor office audited each organization for which it was cognizant about once every 12 years. That audit cycle was far too long because it permitted irregularities to exist for years before they were identified. Under the guidance of the National Science Board's (NSB) Committee on Audit and Oversight, we gradually reduced the audit cycle. We now audit organizations for which we are the cognizant federal audit agency about once every 6 years. The General Accounting Office has indicated that an audit cycle of 3 years is usually optimum, but we believe an audit cycle of 6 years is acceptable in the current budget environment. In addition to auditing financial systems maintained by organizations for which NSF is cognizant, we are required to audit some expenditures under NSF awards at organizations for which we are not cognizant. At these organizations (primarily colleges and universities), we do not review financial system issues applicable to the university as a whole, but we do review particular issues involving individual NSF awards. In 1991, the Senate Committee on Governmental Affairs reviewed our audit coverage of NSF grants at organizations where NSF is not cognizant. Both the Senate Committee and the General Accounting Office recommended that we increase our oversight of NSF grants at these institutions. Despite this direction, our resources are becoming more limited. Reasonable financial projections indicate that, at minimum, we will have to absorb future cost-of-living increases. Some OIG auditors who are now reviewing external awards will have to be reassigned to new internal audits required by the Government Performance and Results Act and the Chief Financial Officer (CFO) Act. In addition, we will have to reduce the amount we spend on external contractor auditors so we can expand contract audit work on NSF's financial statements. In FY 1997, we expect to use approximately $350,000 (about 15 percent) of our audit resources to support the cost of auditing NSF's financial statements. By way of comparison, in FY 1994, we spent $174,000 to comply with our audit obligations under the CFO Act. As audit resources become more limited, we are concerned about our ability to adequately review NSF awards. We are not able to effectively audit all awards that might benefit from a financial review. Accordingly, we modified our audit plan to focus on those awards that are at the greatest risk of misusing funds. To avoid an extension in the audit cycle for cognizant organizations beyond 6 years, we also revised the methods we had been using to decide which cognizant organizations to audit. We believe this will help us keep increases in the audit cycle as small as possible. Our strategic plan for audits explains how we intend to use our audit resources. The plan is to select organizations and awards for review based on a preliminary assessment of whether these organizations would have difficulty complying with regulations that govern the use of federal funds. By using risk assessment principles, we try to identify those organizations and awards that have the greatest risk for financial irregularities and provide for the greatest dollar recoveries. To obtain the maximum benefits from our limited resources, we will continue developing methods of identifying "high-risk" awards for audit. We are using this approach to identify potential auditees out of all organizations that receive NSF funding, including those organizations for which NSF is cognizant. At institutions for which we are cognizant, we will continue to conduct limited reviews of financial and accounting systems on a regular cycle, but we will decide whether full audits are appropriate and determine the scope of the audit based on our assessment of risk. At all other organizations, we will conduct audits only when we identify a specific issue involving a particular NSF award. When we develop audit plans, we first conduct a risk assessment using information that NSF already has available to determine which organizations and awards are at the greatest risk of misusing federal funds. We then conduct in-house surveys by reviewing information obtained directly from potentially high-risk organizations. Information we use to make audit decisions includes findings from prior reviews, patterns of expenditures under awards by cost category, and the organization's internal control procedures. If, after we review this information, we believe the organization is at risk of misusing funds, we conduct either an on-site review or a complete audit. We conduct an on-site review when we believe only a site visit will provide us with the information we need. On-site reviews involve a more in-depth review of the organization's operation than an in-house survey provides. We conduct a complete audit if the information gathered during the risk assessment leads us to believe that funds are at substantial risk. Limited, on-site reviews may also lead to complete audits. We are more likely to conduct an on-site review or a complete audit at organizations for which we have a comprehensive audit responsibility as the cognizant audit agency. The strategic plan will help us conserve audit funds; however, heavy reliance on surveys involves significant risk to NSF. Because surveys provide only a limited understanding of an organization's fiscal situation, it is likely that some organizations and some awards that would benefit from an audit or on-site review will not be identified. However, in the current budgetary environment, we believe the use of risk assessments and surveys to identify organizations and awards for audit will provide NSF with the greatest possible benefits. In this reporting period, we developed a questionnaire to help us identify and evaluate risks associated with NSF awards at grantee organizations. We used information gathered by this questionnaire to rate NSF awards at 450 grantee organizations according to 4 risk factors: our level of audit responsibility (whether we are cognizant), the grantees' familiarity with managing federal funds, (the program's complexity, and the type of program and institution that received NSF funds. We completed risk assessments for large-scale centers and facilities located at universities and other nonprofit institutions, large awards supported by NSF's Directorate for Education and Human Resources (EHR), and grants provided for modernizing academic facilities. Our risk assessments and in-house surveys indicated that it was not necessary for us to conduct audits of NSF awards at 336 of the 450 grantee organizations we reviewed. Based on the results of our risk assessments and surveys, we are reasonably confident that NSF awards at these organizations are expended in a manner that will adequately safeguard federal funds. Of the 450 organizations rated, we identified high-risk awards at 114 organizations. At these organizations, we conducted in-house surveys to determine whether an audit or on-site review was necessary. These in-house surveys provided information on internal controls and costs incurred under the NSF awards. Of these 114 surveys, 94 are in-process and may result in on-site reviews or audits. The remaining 20 in-house surveys resulted in either short, focused, on-site reviews or full audits. At the end of this reporting period, we had completed 8 on-site reviews and initiated 12 audits--5 of which are complete. By using information developed during our risk assessments and in-house surveys, we shortened the time we spent reviewing financial records and questioned over $300,000 of costs in the eight on-site reviews. The five audits identified over $7.2 million in savings for the government. The results of these reviews are discussed below. ***************************************************************** Questioned Cost A cost resulting from an alleged violation of law, regulation, or the terms and conditions of the grant, cooperative agreement, or other document governing the expenditure of funds. A cost can also be "questioned" because it is not supported by adequate documentation or because funds have been used for a purpose that appears to be unnecessary. ***************************************************************** Review of a Major NSF Facility Identifies Substantial Cost Savings During this reporting period, we reviewed several activities at a major NSF-funded facility that is operated by a nonprofit organization comprising institutional members throughout the United States and Canada. In FY 1994, the organization received $122 million in revenue, which included $102 million from federal agencies. Of that $102 million, NSF contributed $66 million through a cooperative agreement. We identified the following three areas of concern: the amount of, and justification for, management fees the nonprofit organization received from federal sources; the FY 1994 allocation of costs among the various projects at the organization; and the organization's implementation of the Cost Accounting Standards (CAS). Fees The government generally pays fees to for-profit entities because those entities must make a profit on their transactions to remain in business. However, the government sometimes pays fees to nonprofit organizations even though fees cover costs the government usually considers unallowable. The government cannot dictate how organizations spend federal money they receive as fees. In 1960, when NSF first began funding the facility, the nonprofit organization requested that NSF pay it a fee to cover both indirect and unallowable costs. At that time, the nonprofit organization had not yet established an indirect cost pool, and it had no other source of income to cover costs that were unallowable federal expenditures. By 1982, the nonprofit organization had steadily increased the revenue it was receiving from NSF and other federal agencies. At that time, the organization established an indirect cost pool because it was receiving funds from entities other than NSF and it wanted to equitably allocate its indirect costs. In 1982, the nonprofit organization also requested that NSF pay it $75,000 per year to cover its unallowable costs. NSF agreed, and it still pays the organization a flat $75,000-per year fee. Because the organization also receives fees from other federal agencies, it has accumulated a large amount of federal fees. In FY 1994, the nonprofit organization received almost $450,000 in fees from NSF and other federal agencies. In January 1995, the nonprofit organization notified NSF that it wanted to charge a 3-percent fee on all non-NSF funds that pass through NSF as interagency fund transfers. If the current level of interagency fund transfers continues and agencies agree to pay the 3-percent fee, we estimate the nonprofit organization will receive over $500,000 in additional fees from federal agencies each year, giving it a total of $950,000 in fees per year from federal sources. NSF, as the nonprofit organization's cognizant agency, is in the best position to assess whether fees in this amount should be paid. We reviewed the total amount the nonprofit organization received as fees from federal and nonfederal sources and asked the nonprofit organization for its justification for assessing the fees. We were told the organization was using fees to maintain a reserve to pay for three major classes of expenditure: maturing liabilities, equipment expenditures, and scientific research. Maturing Liabilities In FY 1994, the nonprofit organization maintained a $3 million cash reserve. We were told this cash reserve was used to pay bills and other maturing liabilities on time. This reserve fund relieves the financial pressure that can occur if the organization incurs a cost before it receives payment from the government for that expenditure and the organization is not reimbursed before the bill comes due. In fact, the organization typically received most of its federal funds either before it incurred the costs or within 5 days of its request for reimbursement. Even if the organization received no funds in advance of incurring liabilities, we believe a $3-million reserve is approximately double the amount of working capital necessary to cover any temporary delay in the receipt of federal funds. Equipment Expenditures The nonprofit organization claimed it needed to use funds from the cash reserve to purchase equipment. We were also told the organization expected to assess its indirect cost pools for this equipment's depreciation. We question the appropriateness of using federal funds to purchase equipment and later charging a portion of the purchase cost against the indirect cost pool as depreciation. Maintaining a cash balance to purchase equipment and later charging the depreciation of the equipment against the indirect cost pool would, in effect, require that the government pay significantly more than the actual cost of the equipment. Scientific Research and Educational Activities The nonprofit organization told us it uses a portion of its fees to support scientific research and educational activities. We are concerned about this use of fees because it appears to fund research by circumventing NSF's peer review system. We believe the organization's requests to fund scientific and educational activities should be evaluated under NSF's peer review system. NSF's peer reviewers and program staff would then be in a position to assess whether the research this organization proposes to fund is more important to achieving NSF's research goals than proposals awarded directly from NSF program funds using widely respected peer review methods. We question whether this use of fees for non-peer-reviewed research is the most effective use of program funds. This nonprofit organization employs approximately the same number of people as NSF. In FY 1995, Congress appropriated $9,000 for NSF's "unvouchered expenditures" (for example, entertainment expenditures) fund. We believe the non-profit organization should be able to establish a discretionary fund that would operate like NSF's unvouchered expenditures fund. The organization could establish this discretionary fund from nonfederal sources by using some of the $1.8 million in unrestricted contributions it receives from nonfederal entities. We recommended that NSF, as the nonprofit organization's cognizant agency, stop paying any fees to the nonprofit organization and notify other federal agencies of its action. If NSF adopts this recommendation, and other federal agencies follow suit, the government would save over $4.5 million over the 5-year term of the cooperative agreement (about $900,000 annually). If NSF decides not to adopt this recommendation, we recommend that it notify Congress of all fees the government intends to pay the nonprofit organization and receive congressional approval for these payments. Cost Allocation All programs at the nonprofit organization benefit from services provided by the facility's budget office and two libraries. The nonprofit organization charges the NSF-funded facility $1.5 million per year for these services. We recommended that the nonprofit organization distribute the $1.5 million among all programs that benefit from the services. By adopting our recommendations, NSF would have saved about $1 million over the life of the 5-year cooperative agreement (about $200,000 annually). NSF agreed to require that the organization distribute the library charges among all programs, but it did not accept our recommendation to distribute costs for the budget office. In this way, NSF will save only $750,000. Cost Accounting Standards CAS identifies 19 areas that should be addressed in a cost accounting system. CAS requires the submission of an annual disclosure statement that describes an organization's cost accounting practices. This description should include a definition of the methods used to distinguish direct from indirect costs and allocate indirect costs. The use of a disclosure statement is intended to reduce the potential for disagreements about an institution's cost accounting practices. If this organization had submitted a disclosure statement, NSF would have known that costs for the libraries and budget office (discussed above) were allocated solely to the major facility rather than distributed among all of the organization's activities. The organization did not want to comply with CAS because it felt the costs outweighed the benefits. In particular, it felt that developing a disclosure statement would be too costly and time consuming. The Cost Accounting Standards Board has analyzed these kinds of concerns and concluded that the initial costs associated with complying with CAS will be offset by reductions in administrative costs associated with preparing cost accounting data to be presented to cognizant federal agencies. NSF management is evaluating our recommendations to eliminate the payment of fees to the nonprofit organization and require that the organization comply with CAS. We will discuss NSF's response in our next semiannual report. Cost Savings Associated With Antarctic Research Vessel In our continuing review of the United States Antarctic Program, we evaluated the management of a subcontract to operate a research vessel (the Vessel). NSF has a contract with a for-profit company (the Company) to administer logistical support functions, such as the operation and maintenance of U.S. facilities in Antarctica and the management of research vessels. The Company has a subcontract with the Vessel's owner to operate the Vessel. The Vessel is an ice-breaking ship that was designed to operate year-round in Antarctic waters and provide a multidisciplinary research platform to researchers in the Antarctic. The Company has a 10-year, $79.4 million subcontract with the Vessel's builder. The subcontract went into effect on March 1, 1992. Under the subcontract's terms, the Company pays the Vessel's owner a fixed amount for each day the Company charters the Vessel. On September 29, 1992, the Vessel's owner requested that the Company increase the subcontract's daily rate because it believed NSF required that the Vessel conduct ice operations that exceeded those outlined in the subcontract's representative mission profile. The Vessel's owner believed the ice operations that exceeded those outlined in the representative mission profile increased its cost to maintain the Vessel and the risk of catastrophic ice damage. Although the subcontract does not refer to the representative mission profile directly, the Vessel's owner believed that the company was required to increase the daily rate to compensate for increased costs as a result of the increased ice-breaking activity. The subcontract between the Company and the Vessel's owner was provisionally modified on August 9, 1993, to address these concerns. NSF has not yet approved the provisional modification. At the request of NSF's contracts branch, we reviewed the provisional modification to help NSF determine whether it should be approved. The results of our review follow. Future Savings During our review, we found three areas where the Program could reduce future costs without affecting the Vessel's operations. Implementation of all of our recommendations should save NSF between $1.2 and $2.6 million. The original subcontract provided the Vessel's owner with 12 paid maintenance days for Vessel repairs, dry-docking, and inspections. The Company's proposed modification eliminated the maintenance days from the subcontract, increased the charter daily rate by $1,500 a day, and recommended that the Vessel remain on-hire for inspections required by the U.S. Coast Guard. We recommended that the Company keep the 12 maintenance days that were in the original contract, increase the charter daily rate by $785 (rather than $1,500), and place the Vessel off-hire during required inspections. If NSF adopts these recommendations, we expect NSF to save at least $931,040. We based this amount on the Vessel owner's estimate of time the Vessel will spend in dry dock for repairs. The owner also proposed that the Vessel be kept on-hire in the event of catastrophic, obviously ice-related damage to any 1 of its 10 ship systems. We recommended that the Company eliminate the catastrophic damage clause from the proposed contract modification because NSF would risk significant liability in the event of long-term damage to the Vessel--liability that, under the current contract, is assumed by the Vessel's owner. Although we made no specific projections regarding future cost savings, the potential savings to NSF could be substantial if the Vessel needs extended repairs for catastrophic damage. In calculating the modified charter daily rate, the Vessel's owner assumed a 5-percent inflation rate to calculate future increases; however, this rate is not in line with current or projected inflation rates. We believe the daily rate increase should be adjusted by current inflation rates (that is, the Consumer Price Index), which averaged 2.85 percent from 1992 to 1994. We do not believe the Vessel's owner should assume an inflation rate for the modification that is higher than current expectations. If NSF adopts our recommendation, it could save $266,278 to use for other scientific purposes or logistics support. Unallowable Costs We questioned $646,266. We questioned this amount because we found the following. The Vessel spent 24 of the 365 days on charter in the second subcontract year in dry-dock for scheduled maintenance. We believe the subcontract's terms dictate that the Vessel be placed off-hire for maintenance and repair. As a result, we questioned $420,822. We questioned $110,509 because the Company paid for the fuel that was used to transport the Vessel to the dry dock from port. However, neither the Company nor NSF required that the Vessel be placed in dry-dock. Therefore, the Vessel's owner is responsible for the fuel cost. During the Vessel's initial testing, the Company deferred acceptance tests for acoustic performance. Before the Vessel was dry-docked in July 1993, the Company subcontracted with two firms to conduct the acceptance tests. The subcontract between the Company and the Vessel's owner states that the Vessel's owner must pay for all acceptance tests. The Company paid the two firms $114,935 to conduct the acceptance tests and argued that the costs were incurred to verify whether the Vessel's acoustic condition met the subcontract's technical requirements. We believe it is the Company's role to verify the Vessel owner's testing results, not to conduct tests of its own. Therefore, we questioned $114,935. NSF's management is evaluating our recommendations. We will report on NSF's response in our next semiannual report. Costs for Travel Services for Peer Reviewers Can Be Managed More Efficiently NSF has a contract with a travel agency to provide travel services for NSF employees. NSF also contracts for the services of another travel agency to provide NSF with administrative services for peer reviewers. The services for peer reviewers include arranging travel plans, processing reimbursement vouchers, and preparing for educational and scientific meetings. The contract for travel services for peer reviewers was the subject of our review. The travel agency receives work orders from NSF divisions that detail how travel funds for peer reviewers are to be spent. In 1985 and 1991, NSF awarded the travel agency travel service contracts totaling over $27 million. The travel agency claimed $19.6 million on these contracts. Claimed costs included charges for transportation, per diem for meals and lodging, rental expenses for meeting rooms and equipment, and contractor fees. We identified four changes in operations that, over 5 years, could save NSF over $500,000. We recommended that the travel agency use NSF meeting rooms and facilities rather than renting meeting rooms at hotels ($221,210), no longer provide for on-site meeting services rather than using the contractor's staff ($141,365), provide funds for lodging and meals to local participants ($124,950); and allow participants to obtain their own meals rather than providing for expensive group meals ($24,415). We questioned over $33,000 that was charged to the two contracts. Specifically, we questioned: $16,356 for transportation compensation, equipment rental, and contractor fees that were overcharged as a result of duplicate payments, billing errors, inaccurate calculations, and excessive per diem costs; $15,231 for meal costs, car rental charges, lodging payments to local participants, and claimed costs that exceeded budgeted amounts; and $1,819 for unallowable travel costs. We also recommended, and the contractor agreed, to remit $113,550 of interest earned on federal funds to the U.S. Department of Treasury. In addition, we found that the contractor did not require that travelers submit travel vouchers after trips and did not complete time and attendance sheets for panel meetings. We recommended that NSF require that the contractor correct these practices. NSF agreed with most of our recommendations for future savings. NSF also began evaluating whether it should continue to provide lodging and meals to local participants. NSF will address that issue and the questioned costs during the audit resolution process. Our risk assessments identified repeated problems associated with accounting for participant support costs. We have developed an audit program that focuses on this type of expenditure. NSF's Grant Proposal Guide defines participant support costs as "costs of transportation, per diem, stipends and other related costs for participants or trainees (but not employees) in connection with NSF-sponsored conferences, meetings, symposia, training activities and work-shops." OMB guidance requires that award recipients obtain prior approval from federal awarding agencies to transfer funds allotted for training allowances (direct payments to trainees) to other expense categories. NSF has implemented this OMB requirement through its Grant General Conditions, which requires that award recipients obtain prior written approval from the cognizant NSF program official before they reallocate award funds that are budgeted for participant or trainee support costs. This rule ensures that funds designated for participants at NSF-sponsored meetings are not used for other purposes without NSF's explicit approval. We are conducting our review of participant support costs in two phases. During this reporting period, we completed Phase I, which involved an evaluation of the extent to which awardees followed ACCOUNTING AND MANAGEMENT PROBLEMS INVOLVING PARTICIPANT SUPPORT COSTS NSF's prior approval policies and obtained the program office's prior written approval before they used participant support funds for other expense categories. Phase I identified 18 institutions from prior audits with participant support cost problems as well as 81 academic and 100 nonacademic institutions for additional reviews. We questioned about $600,000 in participant support costs because 50 awardees spent participant support funds for other expense categories without obtaining NSF's prior written approval. Many of these awardees indicated that they were not aware of the requirement to obtain NSF's prior approval to move participant support funds to other cost categories. We also identified problems with NSF's enforcement of the prior approval requirement. Some awardees either received or believed they received verbal approval from cognizant NSF program officials to transfer participant support funds to other cost categories. Also, some prior approval letters from NSF did not specify the amount of funds approved for transfer. This approach allows awardees to transfer all funds initially designated for participant support to other cost categories. We also identified instances where NSF's program officials approved the transfer of funds after the fact. In some cases, the awardee sought the officials' approval after all of the funds had been spent. In several instances, awardees requested after-the-fact approval when they learned of our review. We recommended that NSF management remind all NSF grantees and program officers about the requirement regarding the transfer of participant support funds and reiterate that after-the-fact approvals will not be routinely approved. We also recommended that management ensure that there is an adequate written record to indicate whether the request was approved. And, if the transfer is approved, how much money may be transferred. NSF is reviewing our recommendations. During Phase II of our review, we intend to: determine the extent to which awardees adhere to the federal cost principles and NSF guidelines by charging only allowable and supportable participant support costs to NSF awards, and consider whether awardees have implemented adequate internal controls to avoid fraud or abuse involving payments to participants. We have identified 18 candidates for on-site audits in Phase II. We expect to complete these audits during the next reporting period. STATEWIDE SYSTEMIC INITIATIVE PROGRAM REQUIRES FURTHER REVIEW NSF provides Statewide Systemic Initiative (SSI) awards to 20 states to improve science education in those states. The SSI program was designed to increase students' knowledge of science and mathematics and give students an opportunity to acquire critical thinking skills. The program is expected to account for approximately $250 million of NSF's budget over a 10-year period. NSF expects significant participation by the states, including management of the awards; cost sharing; and allocation of personnel with scientific, mathematical, and financial expertise. Our risk assessment process identified the SSI program as one that might benefit from further review. We initiated surveys of SSI awards in eight states. Based on the results of the surveys, we began audits in seven of those states. Preliminary audit findings indicate that some problems may exist in the program. We believe an in-depth audit of parts of the SSI program would be helpful to SSI recipients and NSF management. The implementation of the program differs in each state, as does the primary recipient of NSF funding. Primary recipients include state agencies, universities, school districts, and nonprofit professional education organizations. We found that many primary recipients issued subgrants to other subsidiary organizations. It was common for NSF funds to flow through three or four levels of organizations. For example, in a state where the primary grantee was a state agency, the state agency provided a subcontract to a state university. The state university then subcontracted some of the funds to a school district that, in turn, provided some funds to individual schools. In some states, there were as many as 200 implementing institutions. Our preliminary audit results indicated that some primary award recipients were not monitoring costs claimed by subsidiary organizations. We believe this occurred because there were multiple levels of administration, and primary award recipients were confused about their role in the awards' fiscal management. We are determining whether NSF management needs to take additional action to ensure the accuracy of costs claimed by subsidiary organizations. Our initial reviews of records maintained by primary award recipients indicated that support for cost sharing promised by awardees was not always maintained, claimed participant support costs were not documented, award recipients' changes in the use of participant support funds were not approved, and recipients overcharged and mischarged indirect costs. We believe these preliminary findings support our decision to proceed with an in-depth analysis of the SSI program. CHIEF FINANCIAL OFFICER AUDIT ACTIVITIES In Semiannual Report Number 12 (page 7), we reported on the changes to the CFO Act of 1990 that were instituted by the Government Management Reform Act (GMRA) of 1994. These Acts require that we increase the scope of our annual audit of NSF's financial operations. The CFO Act requires that we conduct an annual audit of NSF's Donations Account financial statements, which represent less than 1.5 percent of NSF's $3 billion budget. GMRA expands the audit mandate to require annual audits of all agency funds. The first expanded audit, which involves FY 1996 expenditures, must by completed by March 1997. We are preparing for the FY 1996 agency-wide audit by conducting multiphase reviews in key operational areas that gather the data that will be used in developing agency-wide financial statements. In this reporting period, we began the following. An audit of NSF's 1995 Donations Account financial statements. The second phase of our audit of NSF's general ledger, which includes a review of the daily maintenance of the general ledger structure and account balances. The second phase of our review of electronic data processing controls in NSF's financial accounting system. We are identifying the procedures and controls in various subsystems (payroll, purchasing, disbursing) that provide information to, and receive information from, NSF's financial accounting system. An audit of NSF's budget execution system. We are examining the procedures and internal controls used to execute and monitor NSF's budget. We also contracted with an independent public accounting firm to review NSF's Federal Cash Transactions Report system, including the procedures NSF uses to issue grant funds, record grant expenses in the financial accounting system, and reconcile NSF and grantee cash accounts. We are continuing to participate in discussions with OMB, NSF management, and the NSB to resolve financing issues related to these agency-wide audits. The NSB has approved a $500,000 funding level for the 1996 audit. For the FY 1997 audit, the NSB accepted the Deputy Director's recommendation that the cost for contract services for the financial statement audit be charged to the accounts audited. However, the NSB specified that the amount charged to program and other agency accounts will not exceed $600,000. NSF's CFO awarded a multi-year contract to a public accounting firm for approximately $600,000 to provide a review of the financial management system's auditability and internal controls, as well as recommendations for streamlining and improving the system, formats for financial statements, and the CFO's Annual Report. Once the CFO determines the basic format and content for the agency-wide statements, we will be able to make more detailed plans about our audit approach. OTHER SIGNIFICANT REVIEWS Although the awards we audit involve different projects, our findings often involve similar issues. For example, we often find that grantees cannot obtain and/or account for promised cost sharing. We also find that consultants' and grantees' payments are often excessive or unsupported. Examples of these reviews follow. Audit of a Learning Laboratory for Children Raises Significant Concerns We reviewed two grants totaling over $500,000 awarded to a nonprofit organization to develop, produce, and field test four "learning lab units," and to support a conference on learning in children's museums. The proposed learning lab units consisted of teachers' guides, student workbooks, hands-on exhibits, and software that assist fourth through sixth grade students in informal learning with self-directed, interesting materials. We were concerned because the awardee could not give us schedules of claimed costs, the quarterly expenditures the grantee reported to NSF did not agree with expenditure reports the organization's financial managers provided, and the organization maintained cash balances that were substantially greater than its cash needs under the award. The organization could not provide any documentation to substantiate over $200,000 in costs it claimed had been incurred under the award. In addition, at the time of our site visit, the organization could not provide evidence that it accomplished essential award objectives. As a result, we concluded that claimed costs exceeded the costs actually expended on the NSF award. Based on our preliminary results, we referred the audit to our investigations section for further analysis. School District Charges Incorrect Labor Costs to Award We reviewed two grants totaling $1,087,006 awarded to an urban school district in the northeast. The school district claimed $1,069,504, and we questioned $161,002. We questioned $119,783 in salaries charged to the award because two employees were paid for more hours than they actually worked on the award. We also questioned $14,938 in unallowable costs because the school district incorrectly charged indirect costs to teacher stipends, and charged personal items to the award. The school district's accounting office did not monitor or maintain records of project cost sharing. The school district proposed over $2.1 million in cost sharing under the two awards but could not support $1.5 million of that amount. In addition, the school district neither submitted a final project report nor notified NSF that a principal investigator (PI) who was considered integral to the award left the school district even though the grant's conditions required that NSF be notified. The school district generally agreed with our findings and recommendations. Computer Services Contractor Claims Unallowable and Unsupported Costs NSF awarded a 4-year, $3 million contract to a computer services contractor to develop and maintain software used by NSF's mainframe computer. The contractor claimed $2,994,659, and we questioned $122,295. We questioned $101,311 because the contractor billed NSF twice for the use of the contractor's computer center and computer terminals and did not record all of the claimed direct costs in its financial records. We questioned an additional $20,955 because the contractor charged labor rates that exceeded the negotiated rates. The contractor did not accept our findings and recommendations. The agency will resolve the findings during the next reporting period. Renting Equipment From a Principal Investigator Results In Questioned Costs NSF granted a nonprofit corporation a 3-year award totaling $641,988 to provide training and research experience in aquatic ecology to 30 teachers each year. The corporation claimed $596,208, and we questioned $58,689. We questioned $35,351 because the corporation claimed costs for renting equipment from the PI on the award. When a less than arms-length relationship exists between the lessor and lessee, grantees can claim only the actual cost of the rented equipment, in this case depreciation. We recommended that the corporation credit the award $11,979 because the corporation billed the award directly for items that should have been included in the indirect cost pool and equitably distributed among all of the corporation's activities. The agency will resolve the findings during the next reporting period. School District Needs to Identify Additional Cost Sharing We conducted a survey of a public school district in the northeast that received $3,538,875 under four grants. NSF granted the public school district the awards to further its efforts in curriculum development and teacher enhancement. The school district claimed $1,760,445, and we questioned $3,055 because the school district paid consultants more than the allowable rate. The school district had promised to share the project costs, but it contributed $254,164 less than promised during the first 2 years of the award. In addition, it did not adequately document its cost sharing efforts. The school district accounted for participant support costs as salaries and wages. The district agreed with our recommendations. Zoological Society Needs to Identify Cost Sharing We conducted a survey of a northeastern zoological society that received a $410,481 grant from NSF. The grant provided the society funds to develop carry-along Zoo-Kits for families to use during zoo visits. The zoo had contributed only 10 percent of $131,089 in promised cost sharing but had expended approximately 50 percent of the grant's funds. The zoological society did not periodically review and update its fringe-benefit rate to reflect its fiscal condition. In addition, we could not determine whether the zoological society complied with NSF's statutory salary limitation and the work performed supported the grant's objectives because the zoological society did not have formal consulting agreements. The society's management was receptive to our recommendations. FOLLOW-UP OF RECOMMENDATIONS FROM PREVIOUS SEMIANNUAL REPORTS Management's Response Concerning the United States Antarctic Program's Continental Operations In Semiannual Report Number 12 (page 2), we discussed a review of the operations of the United States Antarctic Program in New Zealand and at McMurdo and South Pole Stations in Antarctica. As a result of our recommendations, the Office of Polar Programs (OPP) is: recouping associated fuel costs (transportation, fuel, estimated labor, and direct material) when it sells fuel to other nations' Antarctic programs; studying the feasibility of using gray-water at the South Pole and McMurdo stations. Antarctic Support Associates (ASA), the Antarctic contractor, agreed to document its consideration of graywater use when constructing new buildings in the Antarctic; planning a complete review of the United States Antarctic Program medical care (until the completion of the review, OPP instructed ASA to ensure that an employee with appropriate medical training is available during the winter at South Pole and Palmer stations in case the station doctor becomes incapacitated); and ensuring that the toilet facilities on the Antarctic aircraft are private. In addition, ASA agreed to: expand its spill response procedures to include training scientific personnel (OPP instructed NSF and ASA personnel to report deficiencies in spill reporting at field camps when noted and requested that ASA retain files with the names of people involved in non-permitted releases of chemicals), and modify its procurement software in Christchurch to reduce duplication of effort. OPP is also conducting a cost-benefit study to determine the most cost-effective way of using the barracks in Christchurch. We expect to describe the results of this study in our next semiannual report. FOLLOW-UP OF RECOMMENDATIONS FROM PREVIOUS SEMIANNUAL REPORTS Management's Response to Recommendations Concerning A Supercomputing Center In Semiannual Report Number 12 (page 10), we discussed a report that detailed at least $5.5 million in savings if the Division of Advanced Scientific Computing (ASC) adopted our recommendations regarding NSF's cooperative agreement with a large commercial company to operate one of NSF's four supercomputer centers. As a result of our recommendations, ASC: requested that the company submit an additional proposal that describes alternate uses for the funds made available as a result of the termination of the supercomputer center's lease obligations (ASC will then consider whether to provide these funds to the company for these new purposes or to fund other research); took steps to deobligate funds that the company had requested for independent research and development, which NSF does not ordinarily fund; and is negotiating with the company to ensure that NSF does not pay the operating costs of activities that are not funded by NSF. In Semiannual Report Number 12, we also discussed the fact that the company billed NSF about $25,000 for consultant fees at rates of pay that exceeded the maximum rates allowed. In a subsequent review of all consultant fees billed to the NSF cooperative agreement, we questioned $57,595 in excessive consulting fees. ASC is consulting with the Division of Grants and Agreements and the Office of General Counsel (OGC) to decide what action to take regarding a potential exemption from state sales tax on the purchase of equipment by the supercomputer center. We will discuss the results of this consultation in our next semiannual report. INVESTIGATIONS The investigations section is responsible for investigating violations of criminal statutes or regulations involving NSF employees, grantees, contractors, and other individuals conducting business with NSF. The results of these investigations are referred to federal, state, or local authorities for criminal or civil prosecution or to NSF's Office of the Director to initiate administrative sanctions or penalties. EMBEZZLEMENT OR DIVERSION OF NSF GRANT FUNDS We place a high priority on allegations involving embezzlement, diversion of grant or contract funds for personal use, or other illegal use of NSF funds. Deliberate diversion of NSF funds from their intended purpose is a criminal act that can be prosecuted under several statutes. We encourage universities and other grantees to notify NSF of any significant problems relating to the misuse of NSF funds. Early notification of significant problems increases our ability to investigate allegations and take corrective action to protect NSF and its grantees. _______________________________________________________ TABLE 1 INVESTIGATIVE ACTIVITY Active Cases From Prior Reporting Periods 28 New Allegations 24 Total Cases 52 Cases Closed After Preliminary Assessment 6 Cases Closed After Inquiry/Investigation 18 Total Cases Closed 24 Active Cases 28 _______________________________________________________ The following section describes cases involving the diversion of funds. Former NSF Division Director Pleads Guilty to Embezzling Grant Funds In Semiannual Report Number 11 (page 22), we discussed a case we referred to the U.S. Attorney detailing abuses by an education association's senior official on NSF and Department of Education awards for a national science education reform program for secondary schools. The senior official was a former NSF Division Director who approved the original grants to the education association for the reform project. Shortly after the official approved the awards, he accepted a position with the education association as the Director of Research and Development. We conducted an investigation and audit of the education association. We determined that the former NSF Division Director knowingly and intentionally embezzled travel funds from the education association, and we questioned $173,408 related to the science education reform project. The education association reviewed our findings and agreed to refund all of the questioned costs. While the senior official worked at the education association, he was responsible for public relations and disseminating materials related to the education reform project. The individual traveled throughout the country as a representative of the association making presentations about science education reform to state and local school districts and professional educational organizations. The association paid for the individual's travel with grant funds. On many occasions, the official also sought reimbursement for travel expenses directly from the institutions and organizations at which the individual lectured, despite the fact that the association had already financed the travel. When the individual successfully obtained reimbursements, he deposited the checks into a personal account without remitting the funds to the association or notifying the association that a reimbursement had been obtained. During our investigation, the individual resigned from his position at the education association. In June 1995, the individual pled guilty in U.S. District Court to violating 18 U.S.C. 666, Theft or Bribery Concerning Programs Receiving Federal Funds, and admitted to embezzling $19,598 in federal funds that were intended to support official travel expenses related to the grants. Sentencing in this case is scheduled for November 1995. Based on the guilty plea, NSF debarred the individual from receiving future federal funds from any federal agency or participating in federal grants for a 3-year period. Principal Investigator Files False Travel Claims A PI charged his NSF grant for travel to several meetings. The PI obtained duplicate reimbursement for this travel from the organizations sponsoring these meetings, which he deposited in his personal bank account. The PI falsely certified to his university and to several other organizations that no other source was paying for his travel. Auditors at the PI's university discovered these false claims. The PI reimbursed the NSF grant for approximately $5,000 in false travel claims and resigned from his university position. We began an investigation and referred the matter to the Department of Justice (DoJ). The PI entered into a settlement agreement with DoJ in which he agreed not to challenge DoJ's conclusion that he submitted four fraudulent travel claims. He agreed to pay a $20,000 penalty under the civil False Claims Act ($5,000 for each of the four fraudulent claims). Under the settlement's terms, the PI may serve as a senior scientist on federal awards under certain conditions. The PI's financial expenditures must be closely monitored by a third party. Travel paid under federal awards must be exclusively for scientific purposes and cannot involve any personal matter. Whenever the PI uses federal funds to pay for travel, any compensation the PI receives must be used for the grant's purposes and may not be retained by the PI. INVESTIGATIONS OF SMALL BUSINESS INNOVATION RESEARCH AWARDS DoJ is working on several cases involving the Small Business Innovation Research (SBIR) program. Two cases are described below. Small Business Sentenced to 5 Years' Probation In Semiannual Report Number 12 (page 21), we reported that an SBIR company agreed to plead guilty to one felony count and pay a civil settlement of $115,000 for submitting false statements that concealed the submission of identical Phase I SBIR proposals to NSF and the National Aeronautics and Space Administration (NASA). On September 11, 1995, the company was sentenced in U.S. District Court to 5 years' probation and 100 hours of community service for making false statements to NSF. Because the false statements concealed the fact that the company had submitted duplicate proposals and received awards from NSF and NASA for the same research, the company was ordered to develop a plan to ensure that it notifies the government of future duplicate proposals. As part of this plan, the company must submit copies of all proposals and bids for contracts to the U.S. probation office for review during its 5-year probation period. The company was also ordered to pay a special assessment fee. Small Business Obtained Over $200,000 By Filing False Claims In March 1995, NSF auditors were not able to conduct an audit of a west coast company's SBIR Phase II grant because the PI, who was the company's president and sole employee, would not respond to their request to initiate the audit. In addition, the PI had not submitted the required final report on his research activities. We conducted an investigation and found that the PI had conducted research only during the first 3 months of the award, even though the award stipulated that he would work on the project for 24 months. After the first 3 months of research, the PI stopped conducting the research and did not notify NSF of this dramatic change in his level of effort. Nonetheless, the PI obtained the remaining $210,000 in grant funds awarded for the research by completing and submitting invoices and Federal Cash Transaction Reports to NSF that falsely certified that he was expending all funds for scientific research under the grant. The PI did not use these funds to support research under the grant. Instead, the PI used at least $53,900 in NSF funds to finance personal investments and repay personal debts. The PI used other funds for travel and equipment not related to the grant's purpose. We referred our findings to the U.S. Attorney to determine whether the PI violated 18 U.S.C. Sec. 287, False Claims; Sec. 1001, False Statements; and Sec. 1343, Wire Fraud. In addition, if the PI is found liable under the civil False Claims Act, the government may recover triple damages, as well as impose penalties of $10,000 for each false claim. We resolved two allegations where we identified improper expenditures but did not find criminal wrongdoing. Principal Investigator Spent Funds Improperly But Did Not Act Fraudulently We investigated an allegation of misappropriation of NSF funds at a small university in the south. According to the complaint, a professor at the university forged certain documents to use grant funds for purposes that were not authorized by the grant. We concluded that the allegation of forgery lacked substance. The professor had obtained written authorization to act for a colleague. In addition, when the professor signed on behalf of the colleague, the professor signed his name as well, acknowledging that he was signing for the colleague. We did conclude that some grant funds were spent improperly. Under NSF's Research in Undergraduate Institutions (RUI) program, the institution received a $66,000 grant to support undergraduate students. We found that the professor improperly charged the RUI grant over $10,000 for equipment and salary costs that were used for an unrelated research project conducted by the professor and graduate-level students. Of the $10,000 in improper charges, $7,000 was spent on stipends for graduate-level students instead of stipends for the undergraduate students who were supposed to receive support under the RUI program. We also discovered that the professor submitted misleading progress reports to NSF. In these progress reports, the professor was required to identify which students were conducting research while he was receiving NSF funds. In fact, all but one of the students mentioned in the professor's progress reports were supported with other funding. As a result of our recommendations, the university agreed to refund to NSF $10,129, the amount improperly charged to the RUI grant for student stipends, supplies, and equipment. The university also agreed to correct misleading progress reports, implement adequate controls to supervise the professor, and initiate appropriate administrative action. Improper Salary Charged to NSF Contract We received an allegation that an NSF contractor charged salary expenses to an NSF contract for work the contractor performed but was not related to the NSF contract. A review of labor charges found that the contractor had charged the NSF contract for a technical service manager's salary when the manager performed duties that were not related to the NSF contract. After the review, the contractor credited $11,399 in salary charges back to the NSF contract. ASSISTANCE PROVIDED TO LOCAL LAW ENFORCEMENT During this reporting period, we analyzed two cases involving the embezzlement of funds intended for science. In both instances, we determined that neither NSF nor other federal funds were directly involved. In these matters, we worked with local law enforcement authorities, and the results of the investigations were provided to local prosecutors. Even though these cases did not directly involve NSF funds, the thefts involved projects that were partially supported by NSF awards. Also, the subjects of the investigations did administer NSF funds at grantee institutions. Ex-Museum Official Pleads No Contest in Theft From Museum A former deputy director of a west coast museum pleaded no contest in state court to embezzling more than $2 million from the museum and its nonprofit foundation. Under state law, the no contest plea is equivalent to a guilty plea. The investigation revealed that the individual used the money to buy cars, pay off personal loans and credit card purchases, and pay taxes on a relative's property. The deputy director embezzled funds received from public entrance fees to the museum and private donations, funds that he was responsible for administering. The deputy director also acted as the institution's authorized representative on federal grants. Since 1990, the museum foundation received over $2.2 million in federal funds, which included 16 NSF grants. We recommended that NSF, acting as the lead federal agency, initiate debarment proceedings to exclude the individual from participating in federal grants for 3 years based on the state conviction. Former NSF Principal Investigator Indicted for Fraud On July 26, 1995, a PI and the PI's secretary were indicted on state Felony Theft and Conspiracy to Commit Felony Theft charges. In February 1995, a state legislative audit of a southern university found that the PI and the PI's secretary had embezzled money designated for an engineering summer program at the university. In 1988 and 1989, NSF supported the university's engineering summer program with two science education grants totaling $100,000 awarded to the PI. The summer program continued after the NSF grants expired with support from student fees, the university, and corporate sponsors. Between 1990 and 1994, the PI and his secretary deposited $231,185 in checks from corporate sponsors and students who attended the university's summer engineering program into a private bank account. The PI and his secretary then wrote and endorsed checks payable to cash, totaling $186,654, from the embezzled funds. The results of the investigation were provided to the local District Attorney who filed the indictment in state court. ------------------------------------------------------------------- TABLE 2 INVESTIGATIVE STATISTICS New Referrals 4 Referrals From Previous Reporting Period 3 Prosecutorial Declinations 1 Criminal Convictions/Pleas 1 Civil Actions Initiated 2 Civil Complaints From Previous Reporting Period 1 Administrative Actions 4 Investigative Recoveries* $66,257 * Investigative Recoveries comprise civil and criminal judgments, fines, and restitutions as well as specific cost savings for the government. ------------------------------------------------------------------ OVERSIGHT The Office of Oversight focuses on the science-engineering-education-related aspects of NSF operations and programs. It oversees the operations and technical management of the approximately 200 NSF programs that involve about 60,000 proposal and award actions each year. The Office conducts and supervises compliance, operations, and performance reviews of NSF's programs and operations; undertakes inspections and evaluations; and performs special studies. It also handles all allegations of nonfinancial misconduct in science, engineering, and education and is continuing studies on specific issues related to misconduct in science. MISCONDUCT IN SCIENCE AND ENGINEERING ****************************************************************** NSF's Definition of Misconduct In Science and Engineering Fabrication, falsification, plagiarism, or other serious deviation from accepted practices in proposing, carrying out, or reporting results from activities funded by NSF; or retaliation of any kind against a person who reported or provided information about suspected or alleged misconduct and who has not acted in bad faith. ****************************************************************** Key Consideration in Applying NSF's Misconduct Regulation NSF's misconduct regulation contains the phrase "fabrication, falsification, plagiarism, or other serious deviation from accepted practices in proposing, carrying out, or reporting results from activities funded by NSF" (45 C.F.R. Sec. 689.1(a)(1)). We interpret the regulation as empowering NSF to take action against serious violations of the "common law" of the scientific community, that is, the shared standards that enable communities of scientists to function. Fabrication, falsification, and plagiarism are examples of the kinds of acts that are so serious as to ordinarily constitute misconduct, but comparably serious transgressions of other kinds are also included. Policy discussions of misconduct have largely neglected the issue of seriousness. Yet, in our view, seriousness is a key consideration in NSF's definition of misconduct. We have recently tried to direct the attention of institutions that perform investigations to this issue, and we are encouraged that some have responded by giving careful thought to seriousness in their investigation reports. Under NSF's regulation on misconduct in science and engineering, adjudicators must consider the seriousness of an alleged offense at two separate points. Threshold Judgment. Adjudicators must first decide whether the alleged offense was serious as part of determining whether it was a "serious deviation from accepted practices." This is a "threshold judgment" that determines whether an act is misconduct. The question at this point is whether the act crosses the threshold dividing "lesser deviations" from "serious deviations." Violations of the ethical standards of the scientific community that are lesser deviations from accepted practices (that is, that fall short of the threshold of seriousness) are outside the scope of NSF's misconduct regulation. The threshold judgment of whether an act was a serious deviation from accepted practices includes a threshold judgment about intent. The crux of this judgment is whether the level of intent is sufficiently blameworthy that the act can qualify as a serious deviation and hence misconduct. In Semiannual Report Number 9 (page 36), we discussed how to make this judgment and distinguished different levels of intent and the kinds of blame that can be associated with them. Degree Judgment. After the adjudicators conclude that a scientist has committed misconduct, NSF's regulation directs them to next "consider how serious the misconduct was." This is a "degree of seriousness judgment" of conduct that has passed the threshold. It locates that misconduct on a continuum. On the basis of this degree judgment, adjudicators decide on the appropriate action or sanction. Thoughtful assessments of the degree of seriousness by investigating officials familiar with the scientific community can help NSF decide what to do when a scientist commits misconduct. If misconduct is found, making the degree judgment about seriousness may involve a fuller consideration of intent than was necessary for the threshold judgment. Our society believes that a naive, thoughtless, or ignorant wrongdoer should be treated less severely than an experienced, calculating, and knowledgeable one. It is appropriate for investigating officials to cite any facts about a person's intent they deem relevant to their degree judgments of seriousness. However, assessing the degree of seriousness involves more than considering intent. In plagiarism cases, for example, we consider the amount of plagiarized material and the originality of the ideas expressed in the copied passages to be relevant to the seriousness of the misconduct. We have noted two recurrent, interrelated problems in how university investigation reports treat seriousness. First, they sometimes confuse threshold judgments of whether conduct seriously deviates from accepted practices with degree judgments of how far beyond the threshold the conduct falls. This leads them to introduce considerations (for example, regarding the subject's lack of experience) into the threshold judgment that should appropriately be reserved for the degree judgment. Second, when investigating committees believe that a finding of misconduct is not warranted, they sometimes explain their conclusion by making strained arguments about intent instead of forthrightly addressing seriousness. Consider two cases we reported in previous semiannual reports. In the first, an inexperienced scientist admitted that he had sought funding from NSF by misrepresenting work he had already completed as work he had yet to perform. The university investigating committee concluded that this was not misconduct because the scientist's intent had been to use the funds for new work, although that aim had been "poorly expressed." However, when we sought clarification of the university's conclusions, the Provost stated that the scientist's act seriously deviated from accepted practice at his university, and the scientist knew that he had already conducted the research in question. We concluded, and NSF agreed, that this case passed the misconduct threshold. The Provost also made a degree judgment about the case. He cited a variety of facts concerning both the act itself and the intent behind it that mitigated the seriousness of the misconduct. We found his degree judgment persuasive, and so did NSF. We do not believe the investigating committee's factual or ethical conclusions about that case were fundamentally different from our own. We believe the committee's report distorted the factual record and evaded the threshold conclusion that the facts required because the committee wanted to avoid unfairly severe actions against the subject. The Provost, by directly confronting the two necessary judgments about seriousness, did more than the committee to achieve this end as well as uphold and articulate the ethical standards of the scientific community. In the second case, a scientist repeatedly used the text of a methodological description written by two other researchers without attribution to the source. There was no evidence that the omission of a citation or quotation was inadvertent or accidental, in the sense that a word processing error might be. To the contrary, the scientist admitted that she did not wish to rewrite the description in her own words for fear that, because English was not her native language, she would unintentionally distort the method she planned to employ. Although it was clear that the scientist intended to deceive her readers into thinking that the words were her own, an investigating committee at her university concluded that she had not committed misconduct because she had "no intent to deceive." In our view, the core issue in this case was not whether the scientist had an intent to deceive. Rather, it was whether the act was a serious violation of community standards. We would have welcomed the committee's thoughtful consideration of what it was about the act itself (that is, the length of the passage, the role of the passage in the proposal, the community's expectations for inexperienced scientists, or some combination of factors) that made it insufficiently serious to be misconduct. Unfortunately, we were left to develop our assessment of the act without much help from the senior scientists acquainted with the ethical environment in which the subject worked. Making good arguments about these two kinds of seriousness is central to investigating and adjudicating misconduct cases wisely and to articulating the common law of the scientific community. Thinking well about seriousness requires reflecting on the community's ethical standards. There are no formulae for calculating seriousness. Scientists who ask for specific, highly codified rules defining misconduct seem to want standards that can be enforced without judgment. Thus, they seek to omit reference to "vague" terms such as "seriousness" in judging misconduct. In doing so, they unwittingly press for a regime of rules that would not truly reflect the subtlety of the ethical norms actually operative among practicing scientists. We doubt that these rules could be so simple to apply as to prevent poor uses of judgment. We believe it is far preferable to face squarely the necessity for judgments about community standards and to encourage reasoned and responsible exercise of judgment than to pretend that the exercise of judgment can be eliminated from misconduct cases or to covertly exercise judgment in ways that avoid scrutiny. Thoughtful discussions of seriousness, in investigation reports and elsewhere, are a good place to start. CASES LEADING TO INVESTIGATIVE REPORTS SENT TO THE OFFICE OF THE DIRECTOR Plagiarism and Violating the Confidentiality of Peer Review We were informed that a subject had submitted a proposal (proposal 2) that contained considerable text, as well as figures and tables, that was copied from a proposal (proposal 1) that the subject had received for panel review the year before he submitted proposal 2. Proposals 1 and 2 were submitted to the same NSF program office and both were funded. The text that was copied into the subject's proposal retained references to software only available at the firm that submitted proposal 1 and references to figures that were found only in proposal 1. In response to our inquiry, the subject told us that he was unaware of the copying. He had hired an undergraduate student to work with him on the proposal. The student would draft a section and then he would review it, making editorial corrections. They repeated this process, section by section, for the entire proposal. The subject admitted that he had provided the student with proposal 1 and said that he thought this was standard practice but said the student had copied the material from it into his submission without the subject's knowledge. We considered it highly unlikely that the iterative process he described could have resulted in the verbatim copy of text from proposal 1 that appeared in his proposal. We concluded that there was substance to these allegations and, at the university's request, we deferred our investigation until it completed its own. The university accepted the subject's statement that he had given the NSF proposal to the student who committed the plagiarism. It concluded that the subject had committed misconduct in science when he violated the confidentiality of peer review by giving proposal 1 to the student. After reviewing the university's investigation report, we concluded that an on-site investigation by our office was required. The subject told us that he hired the student based on a recommendation from another faculty member and gave proposal 1 to the student as a good example of how such a proposal should be written without considering the confidentiality associated with proposals received for peer review. He said he had paid the student from his university research funds and that they began working on the proposal about 6 weeks before the NSF deadline. He claimed only to have written those parts that did not contain copied material and could not explain how the copied material, which constituted the proposed work, had appeared in the proposal. He said he had not noticed the references to the software in the proposed work and blamed the departmental secretary for the references in his proposal to the figures in proposal 1. The other faculty member denied recommending the student to the subject. According to the university's records, the student was not enrolled when the proposal was written. There were no financial records or timesheets to show that the student was paid for this work. We contacted the student, who said he had returned to his native country before the PI claimed he began working on the proposal. The subject had submitted a proposal to the same NSF program office in competition with proposal 1 the year before proposal 2 was submitted. The earlier proposal was declined. The material copied from proposal 1 into proposal 2 was directly responsive to the reviewers' criticisms of this earlier proposal. We found that the subject was an experienced reviewer for NSF who had participated in 2 review panels and reviewed a total of 35 proposals. We did not find it believable that a senior faculty member who had such extensive reviewer experience and who had submitted several proposals to NSF was unaware of the confidentiality associated with peer review. We concluded that a preponderance of the evidence showed that the subject had knowingly violated the confidentiality of the peer review process and that he alone had willfully plagiarized from the proposal received for peer review into his own to improve his chances of receiving NSF support. We viewed his actions as more serious because he failed to accept responsibility for them. He attempted to blame a student, who was not in the country when the proposal was written, for the copying and a secretary for his proposal's inappropriate references to figures found only in proposal 1. We recommended that the Deputy Director conclude that the subject committed misconduct in science, debar him from receiving federal funds for a period of 3 years, and prohibit him from participating in peer review for a period of 5 years. We also recommended that NSF recover the awarded funds from the university. Misconduct Finding for Human Subjects Violations Three families that had been interviewed under an NSF-funded research project complained to the PI's university that she had not fulfilled her promise to pay them for their participation. When attempts to resolve the complaint were unsuccessful, the PI's department chairman referred the matter to the university for investigation. The university found that the subject had misused funds and equipment, violated requirements for the proper treatment of human subjects, and failed to cooperate with the university's investigations. The university also concluded that there was no evidence that the PI had done the work she had proposed to NSF. The PI moved to another institution before the university's investigations were completed, and therefore the university took no action against her. The university recommended that NSF conduct its own investigation, noting that the PI's failure to cooperate had hindered the university's investigation. A scientist and an investigator from our office conducted an investigation at the PI's new institution that largely reaffirmed the university's findings. We determined that the PI had failed to respond to requests for information made by her university's Institutional Review Board for the Protection of Human Subjects (IRB); failed to pay research participants as promised; and, in one instance, failed to obtain school system approval for research involving secondary school students. We concluded that the PI's failures to comply with regulations for the protection of human subjects, taken together, were a serious deviation from accepted practices and recommended that NSF make a finding of misconduct. The evidence also indicated that the PI's misconduct was part of a pattern of habitual disregard for her obligations under her NSF grant. Instances of this pattern included her decision to use her grant funds without NSF's approval for related studies and not for the research she proposed to NSF, misuse of funds from a bank account reserved for compensating research participants, and failure to secure the return or safekeeping of equipment purchased under the grant that was the property of her university. NSF accepted our conclusions and recommendations. It found that the PI's human subjects violations were misconduct in science; reprimanded her; and, to protect NSF's interests as well as those of human subjects, imposed special conditions on any awards she receives before January 1, 1998. This case illustrates the usefulness of the federal misconduct regulation in helping IRBs protect human subjects from abuse, in those situations in which the IRB is not able to exert its own authority. _________________________________________________________________ TABLE 3 MISCONDUCT CASE ACTIVITY FY 1995 FY 1995 First Half Last Half Active Cases From Prior Period 80 81 Received During Period 27 27 Closed Out During Period 26 32 In-Process at End of Period 81 76 _________________________________________________________________ Proposal Submitted to NSF Contains Material Plagiarized From Four Other NSF Proposals We received an allegation that portions of a proposal submitted by an assistant professor at an eastern college were plagiarized from a proposal that originated at another institution. The subject's proposal, a request for funding through NSF's Instrumentation and Laboratory Improvement program, was not funded. Consistent with NSF's position that awardee institutions bear primary responsibility for preventing and detecting misconduct, we informed the college of the allegation so it could further investigate this matter. The subject informed the investigating committee that he had been given two proposals from the other institution. He admitted that he had incorporated 32 lines of introductory material from one of the proposals. However, he claimed that because of time constraints, he had not rewritten the paragraphs in question. The committee concluded that the subject had not intended to plagiarize the material. We found that the college did not sufficiently investigate how the subject used two proposals in preparing his own. Also, the college uncritically accepted the subject's statement of his lack of intent to plagiarize. We opened our own investigation and found that, in addition to the introductory material from the first proposal, the subject had extensively plagiarized from the second proposal from the other institution. The subject admitted that he had plagiarized major portions of his proposal, including much of the scientifically substantive portions of the proposal. Thereafter, the college reopened its investigation and found that in addition to the two source proposals from the other institution, the subject had plagiarized from two proposals that originated from the subject's department at the college. Although the subject copied only short passages of background material from these two departmental proposals, the committee believed that the use of this material was inappropriate because the subject worked on the proposal alone and did not have permission from the authors of the other proposals to copy any material. Moreover, the committee found that 65 percent of the subject's proposal had been copied from the four source proposals. The committee concluded that the subject had committed plagiarism. We believe the subject's largely verbatim adoption of major substantive portions of two proposals that originated from another institution and lesser portions of two departmental proposals is a serious instance of plagiarism. The evidence, including the subject's description of how he prepared his proposal, supports a finding that he acted knowingly. Therefore, we recommended that NSF make a finding of misconduct, specifically plagiarism; reprimand the subject; and debar him from receiving federal grants for 1 year. We also recommended that NSF prohibit the subject from serving as an NSF reviewer or consultant during his 1-year debarment. This case illustrates the importance of checking the available evidence for possible misconduct beyond the scope of that initially alleged. It is insufficient to rely on the subject's word that the misconduct is limited to the original allegation. Moreover, this case illustrates why a finding of misconduct should not require the subject's admission of intent to deceive. Knowing conduct or gross negligence can be inferred from the nature, extent, and repetition of the subject's actions. Plagiarism Falsely Attributed to Student We received an allegation that a faculty member at a southern university had plagiarized his NSF proposal from an award that another PI had received from another federal agency. Most of the text of the proposal was either substantially similar or identical to the text in the award. We learned that the subject may have received a copy of the award from one of his former students who had worked with the PI on the award because the PI had a practice of providing copies of his funded proposals to members of his research team. In response to our request for information about the allegation, the subject asserted that another of his former students (the student) had written the proposal. According to the subject, the student was terminated before he completed his graduate degree. After his termination, the student allegedly approached the subject and volunteered to write the proposal for a field of research that interested the subject, but in which the subject was not an expert. The subject maintained that his participation in the preparation of the research proposal was minimal, and that he submitted the proposal as his own with the student's permission. We referred this allegation to the university for investigation. As a part of its investigation, the university learned from the student that he knew nothing of the proposal or the award, was unfamiliar with the field represented by the proposal, had never been a graduate student at the author's institution, and was employed in another city at the time the subject said he wrote the proposal. The university determined that the subject committed misconduct in science when he plagiarized almost all of the text from the award and that he misrepresented to NSF that the student had written the proposal. The subject resigned from the university and therefore the university took no further action. We wrote to the subject to provide him an opportunity to respond to the university's investigation. The subject responded by reiterating his story that he had been "duped" into believing that the student had written the proposal. We contacted the student, who reconfirmed what he had told the university during its investigation: that he had not written the proposal and that he knew nothing about this specific field of science. Our investigation verified that the student's evidence was reliable. We concluded that a preponderance of the evidence showed that the subject had committed plagiarism when he copied, and submitted as his own, the work of another and that he had acted knowingly. We forwarded our report to NSF's Deputy Director with a recommendation that she find that the subject had committed misconduct in science. We also recommended that NSF send the subject a letter of reprimand and that he be debarred from receiving any government grant support for 3 years. These recommendations are awaiting NSF's action. Plagiarism of Proposals Received in Confidence We received allegations that two NSF proposals from the same department contained plagiarized material. One person was a co-PI on both proposals, and our inquiry indicated that she was responsible for the alleged plagiarism. In each case, the subject allegedly plagiarized material from proposals that NSF had sent her in confidence for merit review. We deferred our investigation to give the subject's university an opportunity to investigate the allegations. The university found that the subject committed misconduct. It prohibited her from submitting research proposals of any kind or accepting research support for projects in which she was the sole investigator for 1 year, barred her from engaging in peer review of any kind for 2 years, barred her from receiving support for new graduate students for 1 year, froze her salary for 2 years, reprimanded her, and informed her that it would immediately dismiss her if she engaged in further misconduct. We recommended that NSF also make a finding of misconduct. We further recommended that NSF reprimand the subject and either debar her from receiving federal grants for one year or enter into a 1-year voluntary exclusion agreement with her. We also recommended that, for 1 year after the debarment or voluntary exclusion ends, NSF require that, when the subject submits a proposal, she ensure that her department chairperson signs an assurance stating that her proposal does not contain any plagiarized material and certify in writing that she has recently reviewed the definition of misconduct in NSF's Misconduct in Science and Engineering regulation, that she has not committed misconduct in preparing the proposal, and that the proposal has been reviewed as described above. We recommended that NSF prohibit the subject from serving as a mail or panel reviewer or as a member of a Committee of Visitors for 3 years. We believe the source of the plagiarized material, the existence of two separate incidents of plagiarism, and the subject's failure to offer a full and frank explanation of these incidents make this a very serious case. In our view, NSF should take strong action against persons who commit misconduct that involves violation of the integrity of its confidential peer review process. Misrepresentation of Academic Credentials in NSF Proposal An eastern university informed us that it had initiated an investigation into an allegation that a PI had misrepresented his terminal degree in an NSF proposal. The university appointed a committee to investigate the allegation. The committee found that the subject had committed misconduct in science when he claimed to have a Master's degree, which he did not. The university determined that the subject had not received any financial benefit from the award and had successfully completed the proposed work. The subject resigned from the university before the committee completed its investigation. Consequently, the committee recommended no actions in the case. Our investigation agreed with the committee's conclusion. We considered that the award was not for research, and it did not require that the recipient have a Master's degree. We believe the subject might apply for NSF funds again. We recommended that the Deputy Director find that the subject committed misconduct in science, the subject receive a letter of reprimand, and the subject be required to certify to NSF for 3 years on any proposal he submits as a PI or co-PI that information contained in the proposal is correct. These recommendations are awaiting NSF's action. CASES SENT TO THE OFFICE OF THE DIRECTOR IN EARLIER SEMIANNUAL PERIOD Plagiarism in Three Proposals Submitted to NSF In Semiannual Report Number 12 (page 29), we discussed a case in which a PI had plagiarized from a source proposal into three proposals that he submitted to NSF. The amount of material plagiarized was substantial. The Deputy Director found that the subject committed misconduct and debarred him from receiving federal grant funds for 1 year. She also excluded the subject from participating as an NSF panelist, reviewer, advisor, or consultant during the debarment period. Since the incident of plagiarism, the subject has moved to a new institution. Because debarment is a serious and public sanction, we recommended that the Deputy Director inform the PI's new institution that the sanction had been imposed on an individual currently in its employ. However, the Deputy Director decided that informing the new institution about this PI's misconduct would be an unnecessary and punitive additional sanction and so she declined to take this action. Misrepresentations of Publications in Proposals Submitted to NSF In Semiannual Report Number 12 (page 31), we discussed a case in which a PI had submitted a proposal that misrepresented the status of several manuscripts as "submitted" when they had not been. Our investigation revealed that the misrepresentations in the NSF proposal were part of a broader pattern of misrepresentation by the subject. The Deputy Director concluded that the subject's false statements to NSF constituted falsification and a serious deviation from accepted practices. The Deputy Director found that the subject had committed misconduct in science and required that, for the next 3 years, when the subject is named as a PI or co-PI on an NSF proposal, he must certify, and his department chairperson must assure to the best of his or her knowledge, that the proposal does not contain any false statements. This certification and assurance must be made separately and confidentially to the Assistant Inspector General for Oversight. Misrepresentation of Credentials In Semiannual Report Number 12 (page 32), we discussed the case of a PI who submitted a proposal in which he falsely claimed to have a Bachelor of Science degree. NSF's Deputy Director concurred with our recommendations in this case. She found that the subject committed misconduct and sent him a letter of reprimand. In addition, the Deputy Director required that, until September 1, 1996, the subject, when he submits a proposal to NSF, must certify separately and confidentially to the Assistant Inspector General for Oversight that all the information in the proposal is correct to the best of his knowledge. SIGNIFICANT CASES CLOSED IN THIS PERIOD WITH NO INVESTIGATION REPORT TO THE OFFICE OF THE DIRECTOR University Finds No Misconduct in Authorship Dispute A university informed us that it planned to investigate allegations of misconduct by an NSF-supported PI. The PI had collaborated on a research project with a postdoctoral fellow who had visited her laboratory. The most serious allegation was that the PI, without her collaborator's knowledge or consent, had changed the order in which the authors were listed (and thus the credit each author received) on a paper based on their collaborative research that she submitted for publication. We agreed to defer our own investigation and advised the university that, in cases of alleged misconduct, we were concerned about whether the subject deviated from accepted practice and, if she did, whether the deviation was serious. Drawing on its own knowledge of the scientific community and the testimony of a senior scientist respected by both the PI and the postdoctoral fellow, the investigating committee concluded that PIs, when submitting papers for publication that report research done exclusively in their laboratories, have broad discretion in deciding the order of authorship. The committee determined that the PI's failure to notify her collaborator or seek the collaborator's approval did not violate a "commonly accepted practice." The university and OIG accepted the committee's judgment and concluded that no misconduct had occurred. However, the committee opined that the subject's actions, though not misconduct in science, fell "below the standard of conduct [the university] should expect of its faculty" because they did not "foster an environment in which its faculty encourage and assist students and post-doctoral fellows in their academic and professional development." The committee recommended that the dean encourage the PI to behave more appropriately. This case illustrates the role that assessments of seriousness play in misconduct investigations and shows that investigating committees can use such assessments to decide cases. It also illustrates that some actions, though not serious enough to warrant a finding of misconduct by NSF, involve failure to adhere to high ethical standards that should concern officials at the university level. Professional Society Conducts Investigation The president of a professional society informed us that the society had received an allegation of plagiarism in a proposal that resulted in an NSF award to the society. The society asked that we defer our investigation to permit a committee of academic scientists appointed by the society to investigate the allegation. Because the two PIs on the award were officers of the society, we took special precautions to guard against real or perceived conflicts of interest that could damage the credibility of the society's investigation and make it impossible for us to use the investigation's findings. We routinely examine whether the members of investigating committees have relationships with either the subjects or the complainants that would create a conflict of interest. In this case, we also requested detailed information about the committee members' relationship to the society's governing council and the scientists in the executive office. We concluded that there was no reason to doubt the committee's ability or willingness to conduct an unbiased investigation. As a result, we agreed to defer our own investigation. The text that was allegedly plagiarized was originally part of an NSF-funded proposal from a PI at a university for a science education project. The project's director wrote a proposal adapted from the PI's original text to apply for renewed NSF funding. The project director supplied the society's PIs with a copy of the renewal proposal and gave them permission to borrow wording from the text of the renewal proposal in preparing their own submission. They treated this as authorization to use verbatim excerpts from the renewal proposal without attribution, which might not have raised a complaint had the project director been a PI on the society's proposal, as the society's PIs had originally planned. However, they changed their plan, with the project director's concurrence, and made him a consultant instead. The investigating committee found that the society's PIs and the project director should have kept the university's PI better informed about their collaboration and should have cited the source of all passages taken verbatim from the university's renewal proposal. Nonetheless, the society concluded that the actions of the society's PIs did not rise to the level of misconduct, noting that the PIs clearly indicated in their proposal that their project was based on the university's project, knew that the project director had informed the university's PI that the society was developing a related project, had reason to believe they were authorized to use the text in question, and sought the university PI's support for their NSF proposal when they submitted it and eventually secured a letter of endorsement from him. The investigating committee reprimanded the PI who wrote the society's proposal for omitting proper citations to the university's proposal. It also said that "it would have been appropriate and courteous" for the PI and her co-PI to have informed the university's PI that they were submitting a related proposal under the auspices of the society before they actually did so. At the committee's suggestion, both of the society's PIs sent notes of regret to the university's PI. The committee further stated that the project director should have more fully informed his project's PI of his role in the society's project. We accepted the committee's judgment that no one had committed misconduct in science, although we assigned slightly different weight than the committee to the different justifications it gave for this conclusion. This case is significant because it is the first time we have relied on an investigation performed by a professional society. We agree with the committee that some ethical transgressions occurred that were not serious enough to be misconduct and warrant NSF action, but that should be acknowledged and corrected. Intellectual Property Dispute A subject submitted a proposal to NSF that contained acknowledgments for two figures, selected scientific data, and reagents to be used if the proposal were funded, to another scientist (the PI) and the PI's collaborator. It was alleged that the PI had neither given the subject permission to reproduce the figures, which he claimed came from his NSF award, nor provided the data or agreed to provide the reagents. The subject said he had received permission from the PI's collaborator to reproduce the figures and the data, and the collaborator had agreed to provide the reagents if the subject's proposal was funded. The subject identified the PI's collaborator as also being his collaborator; however, the collaborator's work with the subject was independent of his work with the PI. The collaborator confirmed the subject's information. The collaborator said he had asked for, and received, permission from the PI to reproduce the figures in the subject's proposal and had orally relayed that permission to the subject. The collaborator had also provided the data, which were freely available from scientific advertisements and product support literature, and had agreed to provide the reagents to the subject. We concluded that the subject had appropriately acknowledged the sources of the information in his proposal. The case was reduced to a dispute between the PI and his collaborator about whether the PI had given the collaborator permission to use the information. We recommended to the collaborator and PI that obtaining and giving written permission to use such materials could help avoid such disputes in the future. We concluded that this dispute did not rise to the level of misconduct and closed the case. Plagiarism Between Collaborators? A subject submitted a proposal that contained a page and a half of introductory text that was copied from another scientist's unpublished manuscript. The introductory text was not indented, and there was no citation to the manuscript. The other scientist's manuscript had been rejected for publication before he showed it to the subject, who then made suggestions on how to improve it. The work described in that manuscript became the basis for a collaboration between the two. The other scientist was responsible for writing the draft of a new manuscript that described the results of their collaboration. He acknowledged that the subject had requested that the introductory text found in the original manuscript appear in the new manuscript. When the subject wrote his proposal, the new manuscript had not been written. However, based on their collaboration, he felt free to use in his proposal material from the rejected manuscript that he expected to be in the new manuscript. At our request, the subject amended his proposal to include a citation to the rejected manuscript and offset the copied text in the proposal. We were also informed that one of the studies described in the subject's proposal was already completed. The complainant could provide no solid evidence to support the allegation. The subject informed us that he had completed a pilot study, and he considered it preliminary data for the full study described in the proposal. It was his understanding that NSF encouraged PIs to include discussions of preliminary data in their proposals. We agreed with the subject. We concluded that the subject's copying text that he presumed would be in a manuscript he was co-authoring and his discussing preliminary data in his proposal were not misconduct in science. We concluded that the two scientists had begun a collaboration that had evolved into a turf battle between the two. In a healthy collaboration, these allegations would not have arisen. STAFF ACTIVITIES Oversight scientists and engineers presented two invited papers: one on federal agency definitions of misconduct in science to the Department of Health and Human Service's Commission on Research Integrity at its June meeting in Chantilly, Virginia, and the other on NSF's response to misconduct in science to an April Colloquium on Research Integrity in Montreal, Canada. In addition, the scientific staff organized a session and presented a paper on how NSF handles allegations of misconduct in science at the annual meeting of the American Sociological Association in Washington, D.C.; delivered a paper on how the federal government deals with misconduct in science at the annual meeting of the Law and Society Association in Toronto, Canada; and presented a paper on the same subject at Amherst College. They also spoke on OIG's inspections and evaluations program at the annual meeting of representatives from institutions served by the Independent Colleges Office. To familiarize NSF with the OIG, scientists and engineers from the Oversight Office have discussed OIG's organization and activities with eight groups of NSF's professional staff in research and education and have served as senior resource staff at an NSF Program Management Seminar for new program managers and other new professional staff. Through workshops, lectures, and seminars, staff have also participated in efforts to apply the requirements of the Government Performance and Results Act. __________________________________________________________________ TABLE 4 ASSURANCES AND CERTIFICATIONS RECEIVED* Number of Cases Requiring Assurances at End of Period 5 Number of Cases Requiring Certifications at End of Period 8 Assurances Received During This Period 1 Certifications Received During This Period 2 * NSF accompanies some findings of misconduct in science with a certification and/or an assurance requirement. For a specified period, the subject must confidentially submit to the Assistant Inspector General for Oversight a personal certification and/or an institutional assurance that any newly submitted NSF proposal does not contain anything that violates NSF's regulation on misconduct in science and engineering. These certifications and assurances remain in the Office of Inspector General and are not known to, or available to, NSF program officials. __________________________________________________________________ OVERSIGHT OF NSF PROGRAMS During this reporting period, we conducted a review to determine whether a limited-term employee had improperly made funding decisions despite a conflict of interest. The results of that review follow. Limited-Term Employee Violated Conflict Of Interest Regulations In March 1995, we received an allegation that a limited-term employee under the Intergovernmental Personnel Act (IPA) at NSF participated in decisions involving his home institution. The IPA worked for a science education program. When the IPA joined NSF in June 1994, he submitted a letter to the director of his science program recusing himself from any matters involving his home institution. In September 1994, the Designated Agency Ethics Official (DAEO) issued a letter to the IPA reiterating that he was "disqualified from participating--by decision, approval, disapproval, recommendation, or rendering of advice--in any matter involving" his home institution. The IPA evaluated science proposals recommended for funding by peer reviewers, including two from his home institution. In managing the evaluations of the proposals, the IPA made recommendations to cut funding for some of the proposals but recommended that one of the proposals from his home institution receive full funding. We determined that, although the IPA avoided taking part in deciding whether to initially fund the two proposals from his home institution, he did participate personally and substantially in the process for deciding on the funding level for these proposals. As a result of our findings, NSF reevaluated the funding levels for all of the recommended proposals in the program. In addition, the Deputy Director issued a letter of reprimand to the IPA and required that he meet with the DAEO for a counseling session. The Deputy Director also informed the individual that his IPA agreement would not be extended. Further, the individual is not permitted to participate in a funded capacity as a PI, co-PI, or equivalent, on the awards in question. We also recommended certain changes in the system NSF uses to identify potential conflicts IPAs may have and ensure that visiting scientists comply with conflicts rules. INSPECTIONS AND EVALUATIONS Our inspections are on-site reviews conducted at NSF or at organizations that receive NSF funding. Inspections' findings and recommendations highlight what works well and identify problems or deficiencies so that managers at NSF and the funded organization can improve their operations and better achieve research and education goals. Inspection teams look for early indications of financial, administrative, or compliance problems so they can be addressed before they become so serious that their resolution requires an audit or investigation. We designed our inspections program to improve our understanding of NSF's grantee activities by integrating financial, administrative, and program analyses in a single review. We view inspections as an effective approach because they allow us to determine whether NSF's program goals are being achieved as well as review the financial and administrative management of NSF awards. Inspections are conducted by multidisciplinary review teams that may include scientists, engineers, auditors, computer specialists, investigators, lawyers, and management/program analysts. During this reporting period, we conducted inspections at a state university and at a primarily undergraduate, liberal arts college. Summaries of the results of those inspections and institution responses follow. Inspection at a State University in the Southwest We reviewed nine grants that NSF's EHR awarded to a state university located in an urban area that has a population of over 1.2 million and within a county where minorities represent over 57 percent of the population. The Directorate awarded the university these grants to improve the science, mathematics, engineering, and technology education of undergraduate students and some were specifically aimed to improve minority students' access to research and research careers. The university's accounting records for cost sharing on equipment purchases were notably good. The university summarized the cost sharing for equipment in separate accounts for each grant and clearly supported its portion of the costs with vendor invoices. While the university generally complied with NSF's and other federal award requirements, we were concerned with the university's application of federal cost principles regarding pay rates for intra-university consulting. The university charged one of our grants for two faculty members' salaries at a rate that was double their regular pay rate. The university did not agree with our finding that intra-university consulting rates cannot exceed the regular pay rate. NSF's Division of Contracts, Policy, and Oversight agreed with our interpretation of OMB Circular A-21 regarding this finding but also acknowledged to the university that NSF's Grant Policy Manual did appear to equate intra-university consulting with external consultants for which a higher pay rate is allowed. NSF has revised the Grant Policy Manual and removed the confusing language. As a result of our recommendations, the university agreed to ensure that its revised policies and procedures for handling allegations of misconduct in science and engineering will meet the requirements of NSF's misconduct regulation, its rules of evidence will be clarified, and that all faculty members and students will be better informed about the university's policies and procedures. Also, the university reported that it is developing policies and procedures that will provide appropriate and necessary access to, and safekeeping of, all research data generated by projects funded by external organizations. Inspection at a Primarily Undergraduate Northeastern College This inspection included 13 awards from 4 NSF directorates and 1 staff office in the Office of the Director. Five of the awards involved the development and/or dissemination of curricular innovations, four supported faculty members' research, two recognized outstanding faculty achievement, one supported a summer program to offer Research Experiences for Undergraduates, and one was for the renovation of the College's chemistry laboratory. A major focus of the program portion of our inspection was on how well NSF serves the needs of scientists and science students at liberal arts colleges. College faculty members and NSF program officers agreed on the criteria NSF should apply in deciding whether to support research proposals from primarily undergraduate, liberal arts colleges. They affirmed that such research proposals should only be supported when the research is comparable in quality to other funded proposals from larger, research-oriented universities. Our review indicated that NSF program officers were receptive to strong research proposals from liberal arts colleges, but that not all College faculty members were confident of this. Our review of NSF projects that supported innovation in undergraduate education examined ways NSF could enhance its efforts to foster communities of scientists who build on one another's work in curriculum development. We also learned that NSF's administration is uneven for awards from cross-disciplinary programs, including awards that recognize outstanding achievement (this inspection included a Presidential Young Investigator Award, Faculty Award for Women, and PI who had received a Visiting Professorship for Women). However, College faculty members who had received outstanding achievement awards from NSF felt that these awards had big effects on their careers. For example, the awards helped faculty members form collaborative relationships with prominent scientists and acquire state of the art equipment that enabled the faculty members to revitalize their research. The College generally complied with NSF's and other federal award requirements. We did find one material instance of noncompliance in that the College did not have time and attendance records to support the faculty members' and research assistants' salaries charged to NSF grants. The College agreed to establish such records but stressed that, despite the absence of these records, it could demonstrate that the work in question was completed and that compensation was within NSF guidelines. The College did not want its failure to complete paperwork to raise questions about individual integrity. The College also agreed to conduct physical inventories of equipment, reconcile equipment records with related financial records, and strengthen and clarify its procedures for handling misconduct in science allegations. LEGAL Attorneys provide legal advice on all OIG activities, including investigations, audits, and oversight of NSF's functions and programs. OIG attorneys supported many of the activities that are described in other sections of this report. Under section 4(A)(2) of the Inspector General Act, OIG is required to review and make recommendations concerning legislation and regulations that affect NSF and NSF-funded activities. OIG attorneys are responsible for conducting these reviews as well as general oversight of NSF's legal activities. Legal Analysis of NSF's Misconduct in Science Regulation NSF's regulation defines misconduct in science as "Fabrication, falsification, plagiarism or other serious deviation from accepted practices in proposing, carrying out, or reporting results from activities funded by NSF. . . ." Although most misconduct cases fall within the categories of fabrication, falsification, or plagiarism, NSF implemented this definition because the agency believed it was necessary to have a general provision such as the other serious deviation clause, because it is impossible to predict the nature of all of the serious unethical conduct that might warrant agency action. The other serious deviation clause sets up a community standard in which the practices of the relevant scientific community determine what is seriously unacceptable conduct. Examples of situations that might be serious enough to violate the other serious deviation clause include tampering with a colleague's experiments, misrepresenting scientific qualifications or achievements in grant proposals, or violating the confidentiality of the peer review process. However, some scientists fear that the clause is too vague and could be used to sanction scientists for conduct that they did not know was prohibited or to punish scientists for using novel or innovative scientific procedures. NSF has supported the clause because it must be able to act against any ethical deviations related to the conduct of NSF-supported research and education activities if they are serious from the standpoint of the scientific community. Therefore, the phrase "serious deviation from accepted practices" is actually the core of the definition of misconduct, whereas fabrication, falsification, and plagiarism are examples of serious deviations. The phrase could not be used to sanction a scientist for novel research or experimentation because it means "deviation from accepted practices" in an ethical sense. Although there have been no findings of misconduct by NSF for the use of novel scientific methodology, and opponents of the clause have found little to complain about the way the clause has actually been used, the controversy over the clause continues. During the debate, the question has arisen whether the other serious deviation clause would pass legal scrutiny if challenged in court. The U.S. Constitution requires that NSF's misconduct in science regulation provide "due process of law." This means that the regulation must sufficiently inform an individual of what conduct is prohibited; if it does not, it is "void for vagueness." However, a legal standard is not void for vagueness just because it does not name every type of behavior that may fall within it; the rule need not state every circumstance that could incur a penalty. Federal courts have invariably found that general standards of conduct, particularly those based on community standards, provide constitutionally sufficient warning of the prohibited conduct. In the leading case in the area of standards of conduct, the Supreme Court upheld the government's standard for dismissal of tenured civil service employees, who "may be removed or suspended without pay only for such cause as will promote the efficiency of the service." The Court held that this general standard was not void for vagueness. It noted that "there are limitations in the English language with respect to being both specific and manageably brief. . . ." The court pointed out that many well-established codes of conduct include broad prohibitions, such as "misconduct," "immorality," or "conduct unbecoming." Courts have consistently held that such broad standards, including prohibitions of misconduct by teachers, professors, attorneys and other licensed practitioners, are not unconstitutionally vague when they are interpreted in light of the practices of the community they regulate. Thus, such broad standards of conduct are appropriate when directed to a specialized group, such as the scientific community, which has its own standards of ethical practice. NSF's procedures for handling misconduct in science cases ensure that an agency finding of misconduct is based on an accurate determination that the conduct seriously deviates from accepted practices in the relevant scientific community (see Semiannual Report Number 10, page 40). The concept of serious deviation, as defined by the relevant scientific community's standard of professional conduct, is no less definite than the standards upheld by the Supreme Court and many other courts. Accordingly, we believe that a court would uphold the other serious deviation clause against a void for vagueness challenge. We are submitting a manuscript for publication that provides a detailed legal analysis of this issue. Reciprocal Consulting Payments Under Education and Human Resources Awards NSF's EHR Directorate sometimes funds large, geographically diverse educational projects by giving several awards to "pilot sites" and one award to an organization to coordinate the pilot site projects. These awards often involve the use of consultants, and we learned that a person working on one project may be retained as a consultant on one of the other projects. We had not previously encountered consulting between recipients of funds from closely related grants, and we did not know whether this practice is widespread. We are unaware of any clear prohibition of the use of federal funds for such consulting (absent fraud), but we believe it is highly susceptible to at least an appearance of impropriety. Therefore, we recommended that EHR assess whether the use of federal funds for consulting agreements of this type among recipients of related federal awards is appropriate, and determine what preventative action it should take if it determines that this type of consulting is inappropriate. EHR reviewed the matter and determined that a consulting arrangement involving an educational association that we had identified was an isolated incident. EHR explained "that EHR strongly believes that the use of federal funds for consulting among recipients of related awards is appropriate when some of the expertise for the success of one award is resident in the project team of the other award, so long as there is no means of coercion applied or conflict of interest, either in fact or perception." EHR has prepared written guidance to accompany award letters for all EHR-funded projects where consultants play a significant role and where grants may be linked under some administrative umbrella. EHR also asked the education association we identified to provide detailed information about consulting expenses to date and required that the association have all future consulting arrangements approved in advance by the NSF program officer. CONFLICTS RULES FOR VISITING SCIENTISTS In Semiannual Report Number 12 (page 43), we discussed the fact that scientists assigned to NSF under the Intergovernmental Personnel Act (IPA) are not covered by the government's financial disclosure requirements or the standards of ethical conduct that apply to federal employees under the Ethics in Government Act. We recommended that NSF develop and implement a contractual agreement that requires that IPAs file financial disclosure forms that are appropriate to their NSF position and comply with the standards of ethical conduct promulgated by the Office of Government Ethics (OGE). We also recommended that the contractual agreement formally notify IPAs that NSF may take disciplinary and/or corrective actions for violations of ethical standards. During this reporting period, the House Committee on Science added a provision to the NSF Authorization Act of 1995 that subjects IPAs to the financial disclosure requirements under the Ethics in Government Act. The bill awaits congressional passage and signature by the President, but we are hopeful that it will become law soon. In the meantime, NSF's OGC has been working with OGE to promulgate NSF regulations that will apply many of the ethical requirements to IPAs. In addition, in response to our recommendations, NSF agreed to require that IPAs file financial disclosure forms and comply with ethics rules, and NSF agreed to notify IPAs that appropriate administrative sanctions will be applied to enforce the ethics and financial disclosure rules. However, by the close of this semiannual reporting period, our recommendations had not been implemented. OTHER LEGAL ISSUES Avoiding Grassroots Lobbying Federal law (18 U.S.C. Sec. 1913) prohibits the use of federal funds by government employees to encourage grassroots lobbying. We learned that an NSF program officer had sent an electronic message that might have been perceived as encouraging some of her grantees to contact their congressional representatives to dissuade them from supporting a pending bill that would have severely cut funding for the program. Although we concluded that the law had not been violated in this instance, we wanted to ensure that, in accordance with OGE regulations, employees were "endeavor[ing] to avoid any actions creating the appearance that they are violating the law. . . ." In the current budget climate, we also felt that it was reasonable to expect that other NSF program officers might be tempted to engage in similar conduct if they were unaware of the prohibitions of section 1913 and the ethics regulations. In response to our recommendations, NSF distributed a legal advisory on grassroots lobbying to all Division Directors and Program Officers. Investigator Financial Disclosure Policy Finalized In Semiannual Report Number 11 (page 40), we discussed NSF's adoption of an investigator financial disclosure policy. This policy requires that an institutional representative certify that the institution has implemented an appropriate conflict of interest policy, all required financial disclosures have been made, and either no actual or potential conflicts exist or all actual or potential conflicts have been managed satisfactorily or disclosed to NSF. During this reporting period, NSF amended its policy to conform with that of the Department of Health and Human Services. NSF revised the policy as follows. Rather than requiring that the PI determine whether a financial interest would "directly and significantly affect the design, conduct or reporting of" NSF-funded activities, the policy now requires the institutional reviewer to make this determination. Conflict management is now required before funds are spent rather than before the award is funded. This will eliminate the need for institutions to review and resolve conflicts for rejected proposals. The dollar threshold for excluded equity interests was increased from $5,000 to $10,000. A certification is now required only from the institutional representative, not from PIs and co-PIs. The policy was extended to apply to large, ongoing projects, such as centers. NSF's Policy was effective on October 1, 1995. Agency Refusal To Provide Information Or Assistance During this reporting period, there were no reports made to the National Science Board of instances where information or assistance, requested under section 5 (a)(5) of the Inspector General Act of 1978, as amended, was unreasonably refused or not provided. Significant Management Decisions That Were Revised No significant management decisions were revised during this reporting period. Inspector General's Disagreement With Significant Management Decisions The Inspector General has no disagreement with significant management decisions made during this reporting period. LIST OF REPORTS NSF and CPA Performed Audits We issued the following audit reports during this reporting period. Where applicable, the total dollar value of questioned costs (including a separate category for the dollar value of unsupported costs) is listed for each report. Date Report Questioned Unsupported Number Grantee Issued Costs Costs 95-1043 National Science Teachers Association 05/30/95 358,428 0 95-1044 Discovery Center of Idaho 04/25/95 0 0 95-1045 H-Tech Laboratories, Inc. 09/29/95 11,821 0 95-1046 Joint Oceanographic Institutions 09/01/95 14,229 0 95-1047 American Society of Mechanical Engineers 09/01/95 65,283 61,767 95-1048 Virginia State Department of Education 09/01/95 317,664 183,316 95-1049 Courtesy Travel Service, Inc. 09/15/95 33,406 0 95-1050 American Management Systems, Inc. 09/29/95 122,263 0 95-1051 Edison Chouest Offshore 09/15/95 646,266 0 INTERNAL AUDITS Audits are designated as Internal when we review NSF operational activities or findings and recommendations of a policy nature resulted from the review and are directed to NSF management. Date Report Questioned Unsupported Number Grantee Issued Costs Costs 95-2105 IPOD/ODP Management Trust Fund Financial Report 05/08/95 0 0 95-2106 Ocean Drilling Program Management Trust Fund Financial Report 05/08/95 0 0 95-2107 Logistics Issues Related to the U.S. Antarctic Program in New Zealand and Antarctica 05/31/95 0 0 95-2108 Conference Contract Procedures 09/18/95 0 0 95-2109 Management Problems Found in Participant Support Reviews 09/29/95 0 0 NSF COGNIZANT AUDITS Date Report Questioned Unsupported Number Grantee Issued Costs Costs 95-4074 American Society for Biochemistry & Molecular Biology 09/05/95 0 0 95-4075 Quality Education for Minorities 09/05/95 0 0 95-4076 Woods Hole Research Center,Inc. 09/05/95 0 0 95-4077 Playing to Win 09/05/95 24,260 8,107 95-4078 Applied Technology Council 09/05/95 0 0 95-4079 California Polytechnic State University Foundation 09/28/95 0 0 95-4080 Nantucket Maria Mitchell Association 09/28/95 0 0 OTHER FEDERAL AUDITS We processed 236 reports provided by other federal agencies. These reports included audit coverage of the NSF programs. The following reports contained questioned costs that required NSF followup. Date Report Questioned Unsupported Number Grantee Issued Costs Costs 95-5503 Elizabeth City State University 04/24/95 7,367 0 95-5544 University of Alaska- Fairbanks 05/26/95 13,639 0 95-5553 Carnegie Mellon University 05/22/95 39,520 0 95-5583 Pennsylvania State University ('92) 05/31/95 4,273 4,273 95-5586 Sacred Heart University 05/31/95 1,586 0 95-5592 Pennsylvania State University ('90) 06/13/95 4,646 4,646 95-5596 Ohio State University 06/13/95 54,066 0 95-5599 University of Hawaii ('93) 06/13/95 31,104 0 95-5602 Stanford University 06/13/95 206 0 95-5639 Stanford University 06/20/95 3,458 0 95-5645 University of Hawaii ('92) 06/21/95 7,297 0 95-5658 State of Arizona 06/27/95 26 0 95-5674 New Jersey Institute of Technology ('92) 08/31/95 688 0 95-5681 Morehouse School of Medicine 08/31/95 275 0 95-5692 Clark University 08/31/95 1,387 0 95-5698 University of Notre Dame 09/01/95 1,374 0 95-5700 Tennessee State Technical Institute 09/01/95 15,994 0 95-5704 State of North Dakota 09/18/95 18,726 0 95-5722 State of South Dakota 09/22/95 113,204 0 SURVEYS Reports in this section are the results of surveys that did not result in a recommendation for full audit but contained questioned costs. Date Report Questioned Unsupported Number Grantee Issued Costs Costs 95-6001 General Atomics 07/14/95 57,595 0 95-6002 Montana Council of Teachers of Mathematics 09/12/95 1,000 0 95-6003 Zoological Society of Philadelphia 09/12/95 996 0 95-6004 Montgomery County Public Schools 09/12/95 3,055 0 95-6005 Canyon Ferry Limnological Institute 09/12/95 58,689 0 95-6006 Maryland Public Television 09/12/95 9,509 0 95-6007 American Association for the Advancement of Science 09/25/95 29,355 0 95-6008 American Indian Science & Engineering Society 09/27/95 27,060 0 95-6009 American Institute of Biological Sciences 09/25/95 4,918 0 95-6010 American Mathematical Society 09/25/95 3,551 0 95-6011 American Psychological Association 09/25/95 18,438 0 95-6012 American Society of Zoologists 09/25/95 700 0 95-6013 Biological Sciences Curriculum Study 09/27/95 1,015 0 95-6014 Cold Spring Harbor Laboratory 09/27/95 39,122 0 95-6015 Computing Research Association 09/27/95 2,150 0 95-6016 Consortium for Mathematics and its Applications 09/25/95 1,335 0 95-6017 Institute for Research on Learning 09/27/95 2,255 0 95-6018 Institute of Global Environment & Society 09/27/95 4,595 0 95-6019 Los Angeles County Museum of Natural History 09/25/95 1,186 0 95-6020 Mathematical Sciences Research Institute 09/25/95 2,418 0 95-6021 National Institute for the Conservation of Cultural Property 09/27/95 8,995 0 95-6022 National Gardening Association 09/25/95 9,494 0 95-6023 National Bureau of Economic Research 09/27/95 580 0 95-6024 Science Learning, Inc. 09/25/95 9,685 0 95-6025 Student Pugwash USA 09/27/95 1,110 0 95-6026 Thames Science Center 09/27/95 44,303 0 95-6027 Boston University 09/27/95 9,256 0 95-6028 Carnegie Mellon University 09/27/95 7,487 0 95-6029 College of Notre Dame 09/27/95 6,317 0 95-6030 Colorado School of Mines 09/27/95 2,779 0 95-6031 University of Connecticut 09/27/95 5,257 0 95-6032 Georgia Technical Research Corporation 09/27/95 2,260 0 95-6033 University of Idaho 09/27/95 2,650 0 95-6034 University of Maryland/ Baltimore County 09/27/95 1,507 0 95-6035 University of Maryland/ College Park 09/27/95 1,084 0 95-6036 North Dakota State University 09/27/95 2,746 0 95-6037 University of North Iowa 09/27/95 1,716 0 95-6038 San Diego State University 09/27/95 496 0 95-6039 University Corporation for Atmospheric Research 09/27/95 10,000 0 95-6040 Rochester City School District 09/28/95 161,002 0 Statistical Information Required by the Inspector General Act of 1978, As Amended Table I. Audit Reports Issued With Questioned Costs Questioned Unsupported Number Costs Costs For which no management decision has been made by the commencement of A. the reporting period. 59 5,363,662 1,793,584 That were issued during B. the reporting period. 68 2,470,122 262,109 Adjustments to questioned costs resulting from C. resolution activities. N/A 345 645 Subtotals of A+B+C 127 7,834,129 2,056,338 For which a management decision was made during the reporting D. period. 52 1,675,600 686,396 (i) dollar value of disallowed costs N/A 712,597 N/A (ii) dollar value of costs no disallowed N/A 963,003 N/A For which no management decision has been made by the end of the E. reporting period. 75 6,158,529 1,374,942 Reports for which no management decision was made within 6 months of issuance. 13 3,767,169 1,112,833 Inspector General Reports ***************************************************************** Funds to be Put to Better Use: Funds the OIG has identified in an audit recommendation that could be used more efficiently by reducing outlays, deobligating program or operational funds, avoiding unnecessary expenditures, or taking other efficiency measures. ****************************************************************** Table II. Audit Reports Issued With Recommendations For Better Use of Funds Dollar Number Value For which no management decision had been made by the commencement A. of the reporting period. 3 4,900,000 Recommendation that were issued during the reporting period. B. (These were issued in two reports.) 10 7,325,000* Subtotal of A+B 13 12,225,000 For which a management decision was made during the reporting C. period. 7 2,300,000 (i) dollar value of recommendations that were agreed to by management 5 1,250,000 based on proposed management action 5 1,250,000 based on proposed legislative action 0 0 (ii) dollar value of recommendations that were no agreed to by management 2 1,050,000 For which no management decision had been made by the end of the D. reporting period. 6 9,925,000 Report for which no management decision was made within 6 months of issuance. 2 4,100,000 * $3.1 million represents recommendations specific to NSF. $4.2 million represents savings for other federal agencies that we identified in our capacity as cognizant auditor for all federal agencies. External Audit Reports With Outstanding Management Decisions This section identifies audit reports involving NSF awards where management had not made a final decision on the corrective action necessary for report resolution within 6 months of the report's issue date. The Division of Contracts, Policy, and Oversight is tasked with making management's decision concerning external audit reports. At the end of the reporting period, there were 13 audit reports with questioned costs that were not resolved. In addition, there were three audit reports with only compliance findings that were open over 6 months. There were 10 reports over 6 months old that required final resolution decisions at the end of the last reporting period. The status of systemic recommendations that involve internal NSF management are described on page 68. ****************************************************************** Management Decision: Management's evaluation of audit findings and recommendations and issuance of a final decision concerning management's response to such findings and recommendations. ****************************************************************** Report Date Report Number Title Issued Reports with questioned costs: 94-1038 Aurora Flight Sciences Corporation 08/08/94 94-1046 Apeldyn Corporation 09/08/94 94-1067 Better Education, Inc. 09/30/94 94-1070 Chemludens 09/30/94 95-1021 Hone Productions 03/01/95 95-1022 BBN Laboratories 03/06/95 95-1023 California BioMarine Products 03/06/95 95-1028 General Atomics 03/27/95 95-1029 BIOSYM Technologies, Inc. 03/27/95 95-1039 Rowe-Deines Instruments, Inc. 03/27/95 95-1040 Innova Laboratories 03/27/95 95-1041 University of Chicago 03/27/95 95-1042 Mr. Wizard Foundation 03/31/95 Reports with only compliance recommendations: 95-1009 American Society of Biochemistry & Molecular Biology 01/04/95 95-5034 Meharry Medical College 02/01/95 95-5052 University of Alabama 02/01/95 For-Profit Claims Unallowed Expenses Period First Reported: April 1, 1994-September 30, 1994 NSF awarded two grants totaling $797,589 to a for-profit organization to develop unmanned aircraft to conduct research on changes in the global climate. The organization claimed $551,626, and we questioned $207,482 because the grantee charged $32,024 for costs incurred under a non-NSF contract; $86,277 in other direct costs that were not approved in the award budgets; and $89,181 in indirect costs that were not applicable to the awards. The organization provided additional documentation, and negotiations are in-process for resolution. We expect these issues to be resolved during the first quarter of FY 1996. For-Profit Corporation Claims Ineligible and Unsupported Costs Period First Reported: April 1, 1994-September 30,1994 NSF awarded a $249,960 grant to a for-profit corporation to investigate coherent fiber-optic communication systems. The grantee claimed $94,883, and we questioned $19,820 in costs that were ineligible for reimbursement or unsupported. We also found that the grantee did not have an adequate financial management system and failed to notify NSF of equipment purchases. NSF and the corporation have agreed on the questioned costs. The corporation promised to provide NSF with documents to support unclaimed costs as offsets to the questioned costs. NSF expects to resolve this audit by December 31, 1995. Grantee Overclaims Indirect Costs Period First Reported: April 1, 1994-September 30, 1994 NSF awarded a $317,800 grant to a for-profit organization to investigate, and provide data on, the benefits of classroom communication systems in various education environments. The grantee claimed $297,550, and we questioned $17,563. We also classified $48,115 in indirect costs as unresolved because the grantee did not have an indirect cost rate proposal prepared for 1993. We also found that cash advances were obtained in excess of cash needs and that property records were not maintained. NSF has not received the data the grantee promised to provide during the last reporting period. Therefore, NSF is taking further steps to resolve this audit. Small Business Claims Indirect Costs as Direct Costs Period First Reported: April 1, 1994-September 30, 1994 NSF awarded a $253,356 grant to a for-profit organization under NSF's SBIR program. The grantee claimed $140,931, and we questioned $72,952 because the grantee claimed indirect costs as direct costs; purchased equipment without NSF's prior written approval; and charged salary costs at a rate higher than proposed in the grant budget. The financial information the grantee provided was not sufficient to resolve the report. We expect final resolution during the first quarter of FY 1996. A Commercial Business Did Not Meet Promised Cost Sharing Period First Reported: October 1, 1994-March 31, 1995 NSF awarded a $347,650 grant to a commercial entity to develop a video-based exhibit to show timed-phased changes in natural phenomena. We questioned $41,600 because the company did not provide cost sharing in the amount it had promised in the grant proposal. We questioned an additional $13,383 because claimed costs were unreasonable and unsupported and interest income on grant payments was not submitted to NSF. In addition, the company failed to submit a property report to NSF. The grantee provided additional documentation to NSF. However, further communication with the grantee is required. NSF expects to resolve this audit during the next reporting period. A Commercial Enterprise Has Questioned Costs Period First Reported: October 1, 1994-March 31, 1995 NSF awarded a commercial entity 12 science education-related awards totaling $11,075,662. We questioned $122,067: $85,404 because the company did not meet cost sharing requirements and $31,605 for indirect costs that exceeded allowable amounts. The grantee has provided NSF data on final indirect cost rates that were established with its federal cognizant agency. In addition, the grantee is preparing information to compare rates charged to NSF awards with indirect costs incurred using final indirect costs rates. The information will be used to determine any offsets to questioned costs. NSF expects to resolve the audit during the next reporting period. SBIR Grantee Claims Excessive Costs Period First Reported: October 1, 1994-March 31, 1995 NSF awarded an $118,879 grant to a small business to develop a controlled environment aquaculture system. We questioned $7,536 because travel costs for family members, alcoholic beverages, and unsupported costs were charged to the award, and indirect costs exceeded allowable amounts. In addition, the company did not submit financial reports or written accounting procedures to NSF when they were due. NSF is reviewing documentation received from the grantee. After further discussions with the grantee, NSF expects to resolve this audit by December 31, 1995. A Supercomputer Awardee Claims Unallowable, Unsupported, and Ineligible Costs Period First Reported: October 1, 1994-March 31, 1995 NSF awarded a $60,895,563 cooperative agreement to a for-profit corporation to manage and operate a Supercomputer Center. We questioned $1,826,288 for independent research and development and capital cost of money charges not allowed in NSF awards; $1,055,634 for unsupported cost overruns; and $51,506 for fringe benefits, indirect costs, and consultant costs charged in excess of allowed amounts. NSF expects to resolve all issues during the next reporting period. SBIR Grantee Has Excessive Indirect Costs Period First Reported: October 1, 1994-March 31, 1995 NSF awarded a $250,000 grant to a small business to develop computer software to be used in building modelproteins. We questioned $83,964 because direct costs were not supported and indirect costs were claimed at amounts that exceeded actual expenses. In addition, the company failed to obtain a commitment for third-party funding, provide required technical reports to NSF on project results, and deposit advanced federal funds in an interest-bearing account. NSF is reviewing information submitted by the grantee, but additionally, a site visit to the grantee may be required. NSF expects to resolve this audit by March 31, 1996. SBIR Grantee Claims Excessive Indirect Costs Period First Reported: October 1, 1994-March 31, 1995 NSF awarded two grants totaling $463,784 to a small business to develop an acoustic doppler directional wave spectrum measurement system and a phased doppler current profiler. We questioned $58,437 because the company claimed unsupported costs and its accounting records and indirect costs exceeded allowable amounts. An additional $61,380 were costs suspended pending NSF's acceptance of the company's indirect cost rate data. In addition, the company did not obtain a third-party funding commitment. The company provided documentation to NSF, but it was not adequate. The company is preparing additional materials to respond to NSF's questions. NSF expects to resolve this audit during the next reporting period. SBIR Grantee Claims Costs Not Approved by NSF Period First Reported: October 1, 1994-March 31, 1995 NSF awarded a $236,574 grant to a small business to develop an instrument for electrical surface characterization. We questioned $23,771 for direct costs and related indirect costs not approved by NSF in the award. The company did not maintain written consultant agreements or segregated duties in the accounting function. Documentation that NSF received from the company is being coordinated with the NSF SBIR program office before a final decision is reached. NSF expects to resolve this audit by December 31, 1995. University Has Overdue Project Reports Period First Reported: October 1, 1994-March 31, 1995 NSF awarded an $11,034,885 cooperative agreement to a university to provide support for a science and technology center aimed at developing opportunities for students and researchers in the astrophysical sciences. We questioned $7,386 because the university claimed the cost of an employee's dental procedure and did not have records to support cost sharing requirements. We also found the university had overdue final project reports for 24 NSF awards and had deficiencies in its property management. Since this audit involves issues that affect other federal agencies, NSF is coordinating with the University's cognizant agency, the U.S. Department of Health and Human Services, for audit resolution. NSF expects to resolve this audit during the next reporting period. Financial Issues Involving Two Principal Investigators Period First Reported: October 1, 1994-March 31, 1995 Two PIs continue to have financial problems with supporting claimed costs similar to those found during a 1990 audit. In 1992, the PIs, as employees of a nonprofit foundation, obtained a new $2,711,043 award to produce films describing exemplary science teaching-methods. We questioned $157,780 because of duplicate first-class air fares, unallowable group lunches, and consultant payments that exceeded allowable limits. We also recommended that NSF restrict to reasonable amounts the salaries and bonus payments the two PIs may charge to the NSF award. NSF has determined that all costs under the NSF award can be charged as direct payments since this is the only work performed by the nonprofit foundation. NSF has prepared communications to the foundation requesting supporting documentation for questioned costs. NSF expects to resolve all issues in the report during the next reporting period. Educational Society Reduces Promised Cost Sharing Period First Reported: October 1, 1994-March 31, 1995 NSF awarded a $375,000 grant to an educational society to support high school chemistry and biology teachers on summer research projects. The society had reduced the promised cost sharing amount for teacher stipends from $600,000 to $450,000 (a 25-percent reduction) without notifying NSF. NSF is reviewing the audit and expects these issues to be resolved during the next reporting period. A Medical College Has Not Met Reporting Requirements Period First Reported: October 1, 1994-March 31, 1995 A medical college that received over $1 million in NSF grant funds did not provide annual progress reports to NSF as required. Also, the college had several internal accounting control problems. NSF has requested that the college respond to the procedural findings in the audit report. NSF expects to resolve this audit during October 1995. University Allows Employees to Charge Retroactive Salary Increases to Federal Awards Period First Reported: October 1, 1994-March 31, 1995 A southern university was found to have allowed an employee to use retroactive pay raises to adjust time and effort reports and charge the costs to federal awards based on the availability of funds. NSF expects to resolve audit issues in October 1995. Additional Performance Measure As required by the Inspector General Act of 1978, we provided in each Semiannual Report to the Congress that give statistical information on work conducted by our audit and investigation units. Tables that provide statistics concerning these required performance measures are on pages 25, 61, and 62. The General Accounting Office, and OMB suggested that Offices of Inspector General develop additional performance measures that provide information about their activities. As a result, we developed an additional performance measure to better explain the work of our office. OIG staff members regularly review NSF's internal operations. These reviews often result in systemic recommendations that are designed to improve the economy and efficiency of NSF operations. We routinely track these systemic recommendations and report to NSF's Director and Deputy Director quarterly about the status of our recommendations. The following table provides statistical information about the status of all systemic recommendations that involve NSF's internal operations. The statistics demonstrate that NSF management has generally agreed to resolve our systemic recommendations in a reasonable manner. Status of Systemic Recommendations That Involve Internal NSF Management Open Recommendations Recommendations Open at the Beginning of the Reporting Period 39 New Recommendations Made During Reporting Period 20 Total Recommendations to be Addressed 59 Management Resolution of Recommendations(1) Recommendations Awaiting Management Resolution 3 Recommendations Resolved by Management 56 Management Agrees to Take Reasonable Action 59 Management Decides No Action is Required 0 Final Action on OIG Recommendation(2) Final Action Completed 22 Recommendations Open at End of Period 37 (1) "Management Resolution" occurs when management completed its evaluation of an OIG recommendation and issues its official response identifying the specific action that will be implemented in response to the recommendation. (2) "Final Action" occurs when management has completed all actions it has decided are appropriate to address an OIG recommendation. Aging of Open Recommendations Awaiting Management Resolution: 0 through 6 Months 3 7 through 12 Months 0 more than 12 Months 0 Awaiting Final Action After Resolution 0 through 6 Months 11 7 through 12 Months 12 13 through 18 Months 3 19 through 24 Months 0 more than 24 Months 8 Recommendations Where Management Decides No Action is Required None to report during this period. Recommendations Awaiting Management Resolution for More Than 12 Months None to report during this period. Recommendations Awaiting Final Action for More Than 24 Months 1. In Report Number, OAO-15-04-88, "Review of the NSF Computer Security Program," April 15, 1988, we recommended that a risk analysis for computer operations be performed and a contingency plan consistent with published guidelines be developed. The responsible office plans to begin preparing the contingency plan in November 1995. Although the delays in implementation are extensive, no losses have resulted. 2. In a July 1, 1992, memorandum, we recommended that NSF have an amendment enacted to bring NSF under the jurisdiction of the Program Fraud Civil Remedies Act. NSF prepared an amendment to the Act and submitted it to Congress on May 11, 1993. Congress did not act on the proposed amendment. The House of Senate staffs have indicated support for the amendment and a willingness to consider including it in suitable legislation when the opportunity arises. NSF management will again include this amendment in NSF's legislative program for FY 1996. 3. In our "Review of NSFNET," dated March 23, 1993, concerning the management and control of the National Science Foundation Network (NSFNET), we recommended that NSF ensure that an audit is conducted of the company that management NSFNET to verify that certain funds have been distributed appropriately, and disseminate the Acceptable Use Policy so that all end-users are award of its prohibitions. The audit is not in progress. NSF has made the Acceptable Use Policy available on the Internet, and will soon publish the Acceptable Use Policy in the Federal Register for public comment. 4. In Semiannual Report Number 8 (page 33), we recommended that NSF issue ethics regulations that address concerns specific to NSF employees. OGC drafted these regulations and they were recently cleared by the OGE. OGC expects to issue these regulations soon. 5. In Report Number, OIG 93-2102, "Review of National Science Board Member's Bank Account," March 31, 1993, we recommended that NSF's Director seek approval from the Comptroller General for funding the awards dinner from NSF's appropriations account. NSF requested an opinion from the Comptroller General. A response is pending. 6. In a September 29, 1993, memorandum, we recommended that all reviewers be more clearly informed, in writing, of all the requirements of NSF's policy regarding the integrity of the confidential peer review process. The goal was to improve the review of proposals submitted to NSF. Language to address this goal has been fully incorporated in revised NSF reviewer forms. These forms are in the clearance process along with other proposed changes to the forms. Prepared by: Office of Inspector General National Science Foundation For additional copies, write: Office of Inspector General 4201 Wilson Boulevard Arlington, VA 22230 For additional information, call: Audit (703) 306-2001 Investigations (703) 306-2001 Oversight (including misconduct in science and investigations) (703) 306-2001 Legal (703) 306-2100 Electronic Mail Hotline: oig @ nsf.gov