DOCUMENT RESUME - ERIC · 2014. 3. 4. · DOCUMENT RESUME. SE 046 608. University Funding: Assessing Federal Funding Mechanisms for University Research. Report to the Chairman, Committee

ED 269 278

TITLE

INSTITUTIONREPORT NOPUB DATENOTEAVAILABLE FROM

PUB TYPE

EDRS PRICEDESCRIPTORS

DOCUMENT RESUME

SE 046 608

University Funding: Assessing Federal FundingMechanisms for University Research. Report to theChairman, Committee on Science and Technology, Houseof Representatives.General Accounting Office, Washington, D.C.G7D/RCED-86-75Feb 8639p.U.S. GeLeral Accounting Office, P.O. Box 6015,Gaithersburg, M1 20877. (First five copies free,additional copies $2.00 each, 25% discount on 100 ormore to a single address).Reports - Evaluative/Feasibility (142)

liF01/PCO2 Plus Poetage.College Science; Educational Finance; Federal Aid;*Financial Support; *Higher Education; *Productivity;Research; *Research and Development Centers;*Research Universities; Science Education; SciencePrograms; Scientific Research; Technology

ABSTRACTAn assessment of the impact that funding mechanisms

had on the productivity and performance of university research ispresented in this report. The General Accounting Office (GAO) studyfocused on five universities that had reputed improvement in programquality. The GAO study concentrated primarily on what funding andother strategies these universities .2sed to improve selecteddepartments and how these departments were able to finance theirimprovement initiatives. Specific questions examined were whetherparticular funding mechanisms played a role in helping universitiesimprove program quality and whether two funding mechanisms, i.e.,individual project grants and center grants had different effects onthe performance of research. It was found that at the fiveuniversities GAO visited, the common element in imp:oves,:nt was anexplicit commitment from the university to improve quality throughincreases in internal ane/or external funding and personnel changes.Initial funding was necessary for building quality, although it camefrom a variety of sources. An appendix contains a summary ofscientists' responses to selected questions on research funding.(ML)

***********************************************************************

Reproductions supplied by EDRS are the best that can be madefrom the original document.

*******************************************************************t***

GAOUnited StatesGeneral Accounting OfficeWashington, D.C. 20548

Resources. Community, andEconomic Development Division

B-221714

rE8 7 1986

Th? Honorable Don FuquaCL arman, Committee on Science and TechnologyHouse of Representatives

Dear Mr. Chairman:

As requested in your November 2, 1984, letter, we have assessed theimpact of funding mechanisms on the productivity and performance ofuniversity research. This report discusses the role particular fundingmechanisms played in 1.1elping universities improve program quality anddifferent effects individual project grants and center grants had on theperformance of research.

We are sendtng copies of this report to appropriate committees of bothHouses, the Director of the Office of Management and Budget, theDirector of the Office of Science and Technology Policy, and the chiefofficials of the following federal agencies: the Departments ofAgriculture, Fnergy, and Defense; the National Aeronautics and SpaceAdministration; the National Institutes of Health; and the NationalScience Foundation. We are also making copies available to interestedorganizations an i iduals.

J. bexter PeDirector

Executive Summary

Over 60 percent of university research funding comes from federalagencies. This research is a key element in the United States' interna-tional competitiveness and technology advancement. Other sources forresearch funding include industry, foundations, and state governments.

Approximately 71 percent of the federal research funds are providedthrough one funding mechanism or category of federal financial supportfor scientific researchindividual project grants. Some scientists andpolicymakers have questioned the consequences of such heavy relianceon individual project grants. For example, does this mechanism dis-courage the performance of innovative, high-risk, and interdisciplinaryresearch?

In response to the House Committee on Science and Technology'srequest that GAO assess the effects of different funding mechanisms onthe productivity and performance of research, GAO looked at:

Whether particular funding mechanisms played a role in helping univer-sities improve program quality.Whether two funding mechanismsindividual project grants and centergrantshad different effects on the performance of research.

In addition, GAO is providing the Committee with a separate report thatdescribes the funding mechanisms used by federal agencies to supportuniversity research and trends in the use of such mechanisms

Background GAO looked at five universities that, according to surveys of the scien-tific community carried out by two education and research organiza-tions, had reputed improvement in program quality. GAO concentratedprimarily on what fumling and other strategies these universities usedto improve the selected departments and how the departments wereable to finance their program improvement initiatives.

Two mechanisms for federal funding of university research are indi-vidual project grants and center grants. Individual project grants sup-port individual researchers who do specific research. Center grants,which account for 9 permit of grants awarded, support broad coherentresearch programs and include coverage of facilities, equipment, andscientific and administrative personnel.

GAO assessed the merits of the two funding mechanisms against four fac-tors that have the potential to affect the performance of research:

Page 24 GAO/RCED416-75 University Funding

Executive Summary

Coverage of resource requirements, which includes trained technicians,equipment, and laboratory space.Stability of financial and resource support, which reflects the continuityand duration of support.Type of research supported, which includes the influence of fundingavailability on the flexibility to pursue new and different areas ofresearch.Administrative burden, which includes researchers' time spent pre-paring proposals, overseeing grants, and reviewing proposals by others.

Results in Brief The particular funding mechanism for university research played alesser role in helping universities improve program quality than theirability to obtain grant funds from such sources as the federal govern-ment, state government, industry, and the university itself.

Responses of scientists to GAO'S questions on coverage of resourcerequirements and administrative burden showed that these factors wereless affected by the particular funding mechanism than by the field ofscience. On the other hand, scientists working under center grantsresponded that they had more stability of financial and resource sup-port and that they were more likely to perform the types of researchdefined as innovative, high risk, or interdisciplinary than scientistsworking under individual project grants.

GAO's Analysis

Improving Research Quality At the five universities GAO visited that were reputed to have improvedprogram quality, the common element in impr wement was an explicitcommitment from the university to improve quality through increases ininternal and/or external funding and personnel changes. Initial fundingwas necessary fcr building quality, although it came from a variety ofsources. Two of the universities received National Science Foundationscience development grants in the late 1960's that enabled them to bringin high-quality junior and senior faculty. Another university receivedstate appropriations that were used to hire new faculty and increase thenumber and quality of postdoctoral fellows. Another university usedfunds from industrial sponsors to implement its plan for programimprovement. (See chapter 2.)

Page 3

5GAO/RCED-86-75 University Funding

Executive Summary

Performance of Research Coverage of resource requirements differed by field of science ratherthan by the type of funding mechanism (individual project or centergrant). Fields of science differ in their needs for such resources as tech-nicians, equipment, and laboratory space. For example, mathematiciansworking on theories may work in isolation with few assistants and littleor no equipment. In contrast, cell biologists may need a number of labassistants, and space scientists may invest large amounts of capital inequipment.

Scientists' concerns about stability of resources and financial er.viron-ment differed depending on their field of science rather than on thefunding mechanism. For example, award duration affects stal, 'litybecause award periods do not always match the actual time needed toperform research. Biochemistry projects may take less time to completethan genetic manipulation experiments in agriculture, where scientistsmust allow a complete new generation of crops to grow before testingcan take place.

Scientists working under center grants reported that they were morelikely to perform types of research defined by the National ScienceFoundation as innovative, high risk, or interdisciplinary than scientistsreceiving individual project grants. For example, 25 out of 32 scientistswith center grants said they proposed research into new areas asopposed to 14 out of 33 scientists receiving individual grants. Scientistsworking under center grants believed they had mere stability andresources to conduct these types of research.

Administrative burden, as measured by the amount of time spent inpreaward activities (applying for awards) and postaward activities(responding to award requirements and reviewing proposals), variedmore by field of science and agency requirements than by type of mech-anism. Defense agency award requirements include postgrant reporting,while civilian agency award requirements include more preawardreviews of proposed research. On the average, scientists in fields, suchas artificial intelligence. that receive awards from defense agencies,reported they spent more time in postaward activities than in preawardactivities. Scientists in fields, such as plant science, that receive awardsfrom civilian agencies reported spending more time in preaward activi-ties. (See chapter 3.)

Recommendations GAO is making no recommendations.

Page 4 GAO/RCED-86-75 University Funding

Eneendve Summary

Agency Comments We did not request agency comments 1.)ecause our work was not carriedout at any agencies and we do not have any adverse comments aboutany agencies or organizations. However, we requested comments on por-tions of the report from the five universities cited in chapter 2 as havingimproved program quality. Those comments are incorporated in thisreport.

Page 5 GAO/PCED -8&75 University Pending

7

Contents

Executive Summary 2

Chapter 1Introduction How the Current Funding System Supports Scientists

Objectives, Scope, and Methodology

889

Chapter 2Role of FundingMechanisms inImproving The Qualityof University Science

Emory UniversityGeorgia Institute of TechnologyUniversity of Alabama in BirminghamUniversity of GeorgiaUniversity of Texas at AustinSummary

14151718192122

Chapter 3Role of FundingMechanisms In thePerformance ofResearch

Coverage of Resource RequirementsStability of Financial and Resource SupportTypes of ResearchAdministrative BurdenSummary

242528313134

Appendixes

Tables

Appendix I: Summary of All Scientists' Responses toSelected Questions

36

Table 1.1: Universities With Reputed Improvement inProgram Quality

Table 1.2: Matched Pairs of UniversitiesTable 2.1: Characteristics of Departments With Improved

Program QualityTable 3.1: FacilitiesTable 3.2: EquipmentTable 3.3: TechniciansTable 3.4: Funding CutsTable 3.5: Changes Over the Last 15 Years in Areas

Affecting Research PerformanceTable 3.6: Funding GapsTable 3.7: Experience With Federal Awards

Page A e?

10

1215

2626272829

3030

GAO/RCED-8615 University Funding

Contents

Table 3.8: Average Time Spent by Scientists in Award-Related Activities

Table 3.9: Shifting Funds

Abbreviations

GAO General Accounting OfficeNIH National Institutes of HealthNSF National Science Foundation

Page 7

33

34

GAO/RCED-86-75 University Funding

Introduction

Since its inception in the late 1940's, the current U.S. system for scien-tific research has emphasized supporting individual scientists' researprojects through national competition for awards. According to theNational Academy of Sciences, the scientific community often associatesthe individual project award system with the success of U.S. basicresrarch and views it as affording the greatest degree of opportunity forpursuit of meritorious ideas.

Despite the belief that the individual project mechanism is closely linkedwith U.S. success in basic research, the House Committee on Science andTechnology has noted problems concerning the current funding systemin which this award type predominates. This report, which wasrequested by the House Committee on Science and Technology, assessesthe roles and impact of different kinds of support for university scien-tific research in different fields of science.

Among the problems with the current system noted by the Committeeand others, such as the National Academy of Sciences, are:

the increased volume of applications for research support that need tobe reviewed;the tendency to fund traditional research ideas rather than innovativeones; andconstraints in the provision of scientific research resources, such asequipment and personnel.

=MM11111.

How the CurrentFunding SystemSupports Scientists

Scientific research in the universities depends heavily on the federalgovernment. In fiscal year 1982 federal agencies provided 64 percent ofthe $7.3 billion spent at universities for research. The federal govern-ment supports university research through a variety of funding mecha-nisms. For purposes of this report, funding mechanisms are categories offederal-financial support for scientific research performed by U.S. uni-versities; they can be divided into direct and indirect support.

Three funding mechanisms directly support research: the individualproject mechanism, program support, and center support. Individualproject awards are typically made to individual scientists for researchthat they have proposed in a discrete research area. This is by far thepredominant mechanism, accounting for 71 percent of agency support.Program support provides support for more than one principal investi-gator in a broad coherent program of research, often multidisciplinaryand long term. Center support provides funding for research projects

Page 8 GAO/RCED-8675 University Funding

10

Chapter 1Introduction

that are coordinated into a coherent program in a broad field of interestat a university. The center award is the only mechanism that providesfunding both for research and for equipment, facilities, and an adminis-trative unit in the university. A recent illustration of the use of thisfunding mechanism is the National Science Foundation's (NsF's) estab-lishment of engineering research centers, designed to strengthen thisfield by providing a concentration of facilities, personne:, andequipment.

Three other funding mechanisms indirectly support research by pro-viding funds for "infrastructure." These funding mechanisms aretraining, equipment and facilities support, and institutional support fora university.

Objectives, Scope, andMethodology

The House Committee on Science and Technology requested that GAOassess the relative merits of different funding mechanisms in terms oftheir effects on the type of research being supported, research perform-ance and productivity, agency procurement administration, manage-ment and administration by the performing organization, and from thepoint of view of the individual scientist. As a result of a literaturereview, the advice of a panel of experts, and consultations with theCommittee, we agreed to assess funding mechanisms as they are used byrecipients in different fields of science at specific research organiza-tions. Our objectives in this assessment were

to determine whether particular fuming mechanisms play a role inhelping universities improve program quality as perceived by the scien-tific community andto examine whether two different types of funding mechanismsindi-vidual project grants and center grants had different impacts on theperformance of research.

Because almost no empirically based literature exists on funding mecha-nisms and their effects on research organizations, we adopted an explor-atory approech to identify those issues that warrant further attentionfrom policymakers. We conducted case studies at 15 different universityresearch organizations. We used two sets of case studies, one focusing onreputed improvement in program quality, and the other on research per-formance and the perspective of individual scientists.

The Committee originally had included research productivity among thefactors it requested we review. However, we determined that we could

Page 9 11 GAO/IRCED-86-76 University Funding


not precisely assess the effects of funding mechanisms on researchquality and productivity because of current limitations in the techniquesfor measuring the outputs of research. Instead, in consultation with theCommittee, we explcred the linkages between the types of supportflowing into research organizations and the reputed research quality ofthose programs.

We focused on how selected university departments were able toimprove their research programs after the federal government hadlargely eliminated special financial assistance for program improvementin the early 1970's. We selected five universities that had successfullyimproved various departments over the past decade on the basis of twonational surveys of U.S. research doctoral programs. The first ("ARating of Graduate Prog -ams") was conducted in 1969 by Kenneth D.Rouse and Charles J. An Jerson for the Ameri,an Council of Education,and the other ("An Assessment of Research Doctoral Programs in theUnited States") was conducted by the Conference Board of AssociatedResearch Councils and published in 1982.

We used the following criteria to select the five universities after con-sulting vi ith the study director of the 1982 survey.

First, where did departments stand in terms of the 1982 survey'sranking of program quality improvcinent as based on responses from

ltists in the same field around the country.Second, which departments showed the greatest change between 1969and 1982 in program reputation, again based on scientists' assessments.

We visited the following universities and departments where we inter-viewed university administrators and faculty members and reviewedprogram improvement documentation and financial records. We lookedat the role of funding mechanisms in the universities' program improve-ment strategies. Due to resource constraints, we focused on departmentsit one geographic regionthe southeastern United States.

Table 1.1: Univers Wes With ReputedImprovement In Program Quality

ii

UniversityEmory University

Georgia Institute of Technology

University of Alabama in Birmingham

University of Texas at Austin

Department/SchoolDepartment of Microbiology and Immunology

School of Chemical Engineering

Department of Physiology and Biophysics

Department of Physics

University of Georgia Department of Botany

Page 10



To respond to the Committee's interest in the effects of differentfunding mechanisms on the performance of research, we designed oursecond set of case studies to explore further sorc. of the problen ; citedwith current federal support for university res' ...rch. Time and resourceconstraints prevented its from assessing all six categories of fundingmechanisms, but the approach we took still sheds light on issuesendemic to all funding mechanisms. Our objective in this second set ofcases was to examine whether two different types of funding mecha-nisms had different imp , on the performance of research. To meetthis second objective, we studied two funding mechanisms, centerfunding and the individual project award mechanism, that together rep-resent 80 percent of the federal dollars obligated for universityresearch. We examined the impact of these two funding mechanisms byexamining four factors related to the perfr --mance of research:

coverage of research resource requirements, which includes trainedtechnicians, equipment, and laboratory space;the stability of support, which reflects the continuity and duration ofsupport;the type r e research supported, which includes the influence of fundingavailability on the flexibility to pursue now and different areas ofresearch; andadministrative bui den, which includes researchers' time spent preparingproposals, overseeing grants, and reviewing proposals by others.

The second set of cases was selected to allow us to examine the use ofmechanisms historically, individually, and in combination at universityresearch organizations. We chose a sample that matched two differenttypos of research organizations (centers and departments), which weassumed would have different experiences with funding mechanisms.We defined centers as research organizations where research projectsare coordinated into a coherent program in a broad field of interest atthe university. Another defining chai;.rteristic of such organizations iscore funding for equipment, facilities, and an administrative unit. WeJoked at centers that had received core funding from a governmentagency for at least 10 years and at departments that had received indi-vidual project awards in that same period of time.

Our sample of matched pairs cut across five fields of science. The finalmatch of departments was made on the basis of location and the degreesto which the department matched the center in terms of types ofresearch done, and other factors, such as seniority of faculty membersand coverage of distinctly different fields of science. The final sample is

Page 11 13 GAO/RCED416-75 University Funding


comprised of 10 of the 25 universities that received the most federalresearch and development support and represents a mix of public andprivate institutions.

Tat le 1.2: Matched Pairs of UniversitiesField of science Center locations Department locationsMathematics University of Wisconsin-

MadisonUniversity if Michigan

Space science

Artificial intelligenceUniversity of Chicago

Massachilsetts Institute ofTechnology

University of Iowa

University of Texas

Cell biology

Plant ,ciencesYale University

Michigan State University

New work University

Cornell University

In selecting different fields of science, we addressed the Committee'sinterest in the impact of different styles of support or combinations offunding mxhanisms on various fields.

Our data collection efforts involved the administration of a structuredquestionnaire to principal investigators at the various universities. Wealso asked universities to provide us With data on their use of differentfunding mechanisms from federal and nonfederal sources :n 1970, 1975,and 1984-85.

The questionnaire was administered to assistant, associate, and fullprofessors at the universities we visited. In all we interviewed 70research faculty. Using this questionnaire, we gathered data on avariety of factors bearing on the perceived impart of federal individualproject grant awards versus federal renter awards in terms of coverageof resources, stability, types of research, and administrative burden.These factors are discussed in detail in chapter 3.

In all cases, data were cross tabulated by type of research organizatie(department or center) and by field of science (artificial intelligence,space science, mathematics, cell biology and plant science). In addition, aseries of open-ended questions were asked to develop additional infer-mation about the perceived effects of funding on scientific research.These questions were designed to create small-scale case studies whenthe comments of all scientist.; in a particular center or department wereaggregated.

Since the case study approach was used to address both objectives, animportant caveat must be noted. lur study is not representative of all

Page 1214

GAO/WED/1646 University Pending


fields of science, the totality of U.S. resew ch universities, or all federalagencies or components of agencies.

We did not request agency cor.iients because our work was not carriedout at any agencies and we do not have any adverse comments aboutany agencies or organizations. However, we requested comments on por-tions of the repot 4- from the five universities cited in chapter 2 as havingimproved program quality. Those comments are incorporated in thisreport.

Page 18 15 GAO/RCED-86-75 University Funding

Chapter 2

Role of Funding Mechanisms in Improving theQuality of University Science

This chapter assesses the role of funding mechanisms in improving theperceived program quality of university science departments. In the1d6O's federal agencies developed several funding mechanisms designedeither to create new research expertise or to increase existing researchexpertise. These funding mechanisms had been discontinued by theearly 1970's. In an effort to determine how selected .iniversity depart-ments ere able to improve their academic and research programs whenthe federal government had eliminated special financial assistance forrepmrch program improvement, we visited five universities thataccording to national surveys had successfully improved variousdepan ments over the past few years. (See objectives, scope, and meth-odology in chapter 1.) This chapter concentrates primarily on whatfunding and other strategies these universities used to improve theselected departments and how the departments were able to financetheir program improvement initiatives.

We found that these departments financed program improvement plansby obtaining funds from federal grants, state government, industry, oruniversity sources. With these funds the departments hired additionalfaculty, renovated research facilities, and purchased new equipment.These actions contributed to the quality of their research programs andenabled the departments to compete successfully for additional externalgrants and contracts. Although the departments used a variety offunding mechanisms, the individual project grant was the principalmechanism used by all the departments. Two departments received spe-cial science development grants from the National Science Foundation inthe mid-1960's. Table 2 1 briefly summarizes the information we foundconcerning these funding mechanisms and program improvement strate-gies for these five departments. More detailed summaries follow thetable.


Chapter 2Role of Funding Mechanisms in Improvingthe Quality of University Science

Table 2.1: Characteristics of Departments With Improved Program Quality

Federal research funds1970 1984

$ 140,466 $1,158,441

Universitynary

University,MicrobiologyandimmunologyDepartment

GeorgiaInstitute ofTechnology,School ofChemicalEngineering

University ofAlabama atBirmingham,PhysiologyandBiophysicsDepartment

Funding sources used to improveprogramUniversity awarded $620,000 todepartment as seed money

PercentChange Key elements of improvement

+725 Seed funding used to increase thenumber of tenured faculty members;new department chairman in 1979

Increase in support kohl industry,federal government, and foundations(industrial sponsors)

149,016 754,273 +406 1978 implementation of written planfor improvement of program. Planfocused on faculty recruitment andimproving university relations withindustry

Seed money from stateappropriations

240,401 2,488,969 +935 New chairman in 1979, focus on hiringnew faculty and increasing thenumber and qu,..'ity of postdoctoralfellows

University of !967 NSF Science Development grant 405,695Georgia, of $972,000 matched by an infusion ofBotany state funds and start-up funds fromDepartment the university for new researchers

Individual research grant sustainsprogram improvern,3nt. unrestrictedincome from an encowment fund

1,673,874 +313 Support through a variety of fundingmechanisms allowed expansion ofspace for taculty and studentresearch and the addition of morefaculty, equipment, graduatestudents, and postdoctoral fellowr

University of 1966 NSF Science DevelopmentTexas at grant. Department st-engthened byAustin, income from private endowmentPhysicsDepartment

1,762,154 7,825,487 +344 Science Development Grant providedthe opportunity to bring in high-quality junior and senior faculty withinitial research supportUniversity funding proceduresenhanced acquisition of equipment.thereby improving program quality.

'111ImulEmory University The Chairman of the Microbiology and Immunology Department told us

that the department began its greatest period of growth and improve-ment in 1979, when he was hired. The chairman described the depart-ment at that time as a modest, but decent one, which he believed couldbe expanded into a well-balanced, nationally recognized, high-qualitydepartment. The university's administration also wanted to improve thequality of the department and agreed to provide about $620,000 in"seed money" to increase the number of tenured faculty. Additionalfunds were provided to acquire more modern equipment for instruc-tional and research purposes and to support additional graduate andpostgraduate students. In addition, the university agreed to renovatespace for the Microbiology and Immunology Department. According to


17


the department chairman, renovation costs were between $1.5 millionand $1.75 million.

The chairman told us that the first priority for improving the depart-ment was to hire additional faculty members who were highly trained,prominent in their field, and who would aggressively seek externalresearch funds through grants and contracts. When the chairman washired in 1979, the department had eight faculty members. Today, thedepartment has 11 faculty members, 5 of whom have been hired sincethe new chairman came on board. The current faculty has successfullyincreased the department's external funding from about $240,000 in1979 to over $1.5 million in 1985, including about $1.4 million in federalfunds. The department would like to increase its faculty to 16 or 18members, but current space constraints have precluded further growth.

Acquiring additional equipment for research and instructional purposeswas another high priority for improving the department. A 1978appraisal of the department's laboratories concluded that existingequipment was not suitable for modern research approaches in microbi-ology . Since then, the department has purchased several new pieces ofequipment.

According to the department chairman, applicants for predoctoral andpostdoctoral training in the department have also increased in numberand quality. In 1979 the Microbiology and Immunology Department hadonly 5 graduate stunents; today it has 24. The department has providedfinancial support for six of the predoctoral and postdoctoral studentsthrough a training grant from the National Institutes of Health (Nix).This grant, which began in July 1984, will provide a total of $499,640over a 5-year period. The university has also increased its student fel-lowship support for this department from $32,500 in 1979 to a 1985level of $65,700 per year.

NIH'S Biomedical Research Support Grant provides additional funds onthe basis of total amount of NIII grant dollars received by Emory. Theuniversity then shares these funds with various departments as theneed arises, for example, to put hase expensive pieces of researchequipment or provide interim support for faculty who are "betweenresearch grants." Research funds from the Multiple Sclerosis Society,the American Cancer Society, the Rockefeller Foundation, and the stateof Georgia provided about $150,000 in 1984, or about 11 percent of thedepartment's external research funds. Because Emory is a private uni-versity, it does not receive an appropriation from the state of Georgia.

Pagot 16

18GAO/RCED-86275 University Funding


Georgia Institute ofTechnology

School of Chemical Engineering officials told us that substantialimprovements that were made in the quality of its faculty, graduate stu-dents, and educational program would not have been possible without aflexible university administration, a determined newly appointed Chem-ical Engineering director, and a supportive faculty. in a time ofdecreasing federal support for program imr vement, Chemical Engi-neering developed a comprehensive written .gin for improving thequality of its program. The essence of its plan was to achieve excellenceby improving the quality of its faculty and graduate students.Improving relations with industry was also a priority.

Since 1978 the Chemical Engineering School has successfully attracted11 new faculty members. The Director of Chemical Engineering, inreflecting on the improvement in quality of the school, cited a number offactors responsible for the successful recruitment of highly qualifiednew faculty. The factors he cited were

a perception that the rapidly changing Georgia Tech Chemical Engi-neering program would be a good place to build or continue a career,the willingness of the Dean of Engineering to permit the school to recruitfaculty at all ranks, andattractiveness as a place to work and live.

Faculty recruitment took priority over building the Chemical Engi-neering graduate enrollment. The depllianent established an initial goalof four to five graduate students per faculty member and carried outextensive recruiting efforts to achieve that goal. In the summer of 1978Chemical Engineering had only 12 graduate students; today, it has :_bout100.

In addition to improving the quality of faculty and graduate students,improving communications and relationships with industry was also apriority of the school director. He believed a good relationship withindustry not only enhances educational opportunities for the students,but also increases industry's financial support for the program and con-tributes to the institution's stature. Activities aimed at improving theschool's external relationships, including industry, during the past fewyears included

establishing external advisory boards comprised of industrial and aca-demic represeqtatives interested in the program,publishing a .tew graduate program booklet containing specific programinformation and listing the research interests of individual faculty,

Page 17 GAO/t10ED-S6-75 University Funding

19

the Quality of University Science

Chapter 2Role of Funding Meciumisms in Improving

issuing an annual alumni newsletter since 1979, andpursuing opportunities for interaction with industrial representatives.

The budget for Chemical Engineering has increased dra.71atically duringthe past 15 years. In 1970 the budget was approximately $582,000, butby 1984 the budget had grown to more than $3.5 million. The greatestbudget increases have occurred since 1978, the year the new directorwas hired.

The increase in funds has come from several sources including the stateof Georgia, the federal government, and industry. Because the ChemicalEngineering School performs extensive research, a substantial part of itsfunds come from grants and contracts from industry and governmentagencies. In 1970 the state of Georgia supplied 68 percent of its funds,with the remaining 32 percent provided by industry, the federal govern-n. ,nt, and foundations. By 1984, however, the trend was away fromstate support, with only 50 percent of the school's funds coming fromthe state. The remaining 50 percent of the $3.5 million budget camefror i such external sources as industry, the federal government, andfoundations (including industrial sponsors).

University of Alabamain Birmingham

Much of the Physiology and Biophysics Department's improvement, asreported in the 1982 "Assessment of Research Doctoral Programs in theUnited States," has occurred since 1979 when a new chairman washired. According to the department chairman, the goal of the univer-sity's administration and departmental faculty was to accelerate themodest expansion that had taken place in previous years and generallyto broaden the scope of research in the department. The departmentemphasized recruiting new faculty, consolidating the faculty into asingle functional unit, purchasing new scientific and word processingequipment, restructuring the graduate program, and starting a series ofdepartmental seminars featuring nationally mognized speakers frunother universities. Of these stated goals, tile chairman told us that thedepartment has been most successful in improving the quality of itsfatuity and increasing the number and quality of its postdoctoral fel-lows. University officials attributed much of the department's improve-ment to a supportive and flexible university administration, asubstantial increase in state funding, and the strong leadership of thenew department chairman. A substantial increase in external funds alsohelped finance the program improvement initiatives.

Page 18 2 0 GAO/RCED-88-75 "Jniversity Funding


The department chairman, in reflecting on the improvement in qualityof the department, stated that his number one priority upon arrivingwas to build a strong research program. He believed this could beachieved by hiring the best possible researchers in their respectivefields. Because of the university's willingness to hit e faculty at all ranksand to pay highly competitive salaries to get them, the department hasbeen successful in attracting 10 researchers since 1979. The chairmandescribed these researchers as outstanding and as having internationalstature in their research field. These faculty members have aggressivelysought external research funds that have helped to support the programimprovement plans.

Funding for the department has grown dramatically over the last 10years. In 1975, for example, the total departmental budget was only$464,880. It had grown to $1.7 million in 1980, but by 1985, the budgethad increased to more than $5.5 million. Department officials eaimatetithat individual project grants make up at least 90 percent of awards intheir department, and that the ability to compete successfully forexternal research money is one key to the program's success. Most of theincreased funding has come from additional federal mon y for research,but substantial increases also occurred in funds from state appropria-tions and from nonfederal health agencies such as the American HeartAssociation, the American Cancer Society, and Cystic FibrosisResearch Center. According to University officials, "seed money" fromthe university's state appropriation helped start the program improve-ment initiatives.

The Pk'siology and Biophysics Department Chairman told us that thedepartment has also been successful in attracting outstanding sTaduateand postgraduate students. The most impressive growth has been in thenumber of postdoctoral fellows. In 1979, for instance, the departmsnthad only seven postdoctoral fellows. By 1984 that number had grown to22, compared with a national average of 6 in a typical phjsiologydepartment.

University of Georgia University officials cited several factors that have been responsible forthe improvement in the Botany Department.

The university was committed to developing an excellent department.In 1967 the university received a $6.0 million NSF Science DevelopmentGrant. The Botany Department's share of the grant was $972,000. Thesefunds and a commitment of funds from the state government enabled

Page 19

21GAO/RCED86-75 University Funding

Chapter 2Role of Funding Mechanisms in Improvingthe Quality of Unive sky Science

the department to increase the faculty size from 15 to more than 20 andto purchase new equipment.The state provided over $3.4 million to build a new 157,000 square footplant sciences building and allocated to the Botany Department 60,000square feet for teaching and research facilities. The new space assistedin the recruitment of desired faculty specialists, and shared space pro-moted interdepartmental cooperation and communication. Fait of thecost of this new building ($500,000) tame from an NSi. Science Develop-ment Grant.The university provides start-up funds for new researchers. Dependingon the area of research, start-up costs range from $15,000 to $100,000per researcher. For example, it costs about $100,000 to set up a plantmolecular biologist with the necessary laboratory facilities and equip-ment to compete for external funding.St:ong leadership from the university administration and BotanyDepartment faculty promoted and encouraged research, which attractedexternal research funds. Federal research funds, for example, grewfrom $41,000 in 1965 to almost $1.7 million in 1984.In more recent years, income from a $1-million endowment fund, desig-nated solely for the Botany Department, has also provided substantialunrestricted money that the department can use for special needs suchas research equipment, student assistance, and travel.

Along with the improvement in faculty, research equipment and facili-ties, `he department chairman believes the quality of graduate studentshas also improved. Currently, the Botany Department has about 50graduate students, about 30 of whom receive teaching assistantshipsand 20 of whom have grant funds.

Although NSF'S Science Development Grant served as a catalyst for pro-grain improvement, university officials believe that the individualresearch grant has been the major funding mechanism that has sus-tained the program improvement momentum. They believe a depart-ment needs start-up or "seed money" to attract high-quality faculty andprovide necessary research space and equipment, but after that, theindividual research grant is the mechanism for achieving the highestquality science research.

The Botany Department :gas experienced remarkable growth in funding.Federal funding has grown from $41,000 in 1965, to $405,000 in 1970(includes part of the NSF Science Development Grant) to almost $1.7 mil-lion in 1984. Total lepartment funds from the state and federal govern-ments, industry and foundations, and endowment income grew from

Page 20 GAO/RCIXO MI5 University Funding

22

Chapter 2Role of Funding hiechanisms in Improvingthe Quality of University Science

$1.7 million in 1980 to more than $3.0 million in 1984. Most of thisgrowth has been in federal research funds through individual researchgrants.

University of Texas atAustin

According to the Physics Department Chairman, since receivir% an NSFScience Development Grant in 1966, the department has made progressin improving the quality and number of faculty and graduate studentsand in improving its overall research program. Funds provided by thegrant were used for (1) additional faculty, (2) initiation of new researchactivities, (3) establishment of a Faculty Associate Program wherebyrecent doctoral recipients were brought to campus for 2-year periods ofintroduction to teaching and research, an i (4) initiation of a program ofcurriculum development. University administrators stated that a majorp:r.itive effect of the NSF Science Devel-pment Grant was the opportu-nity it provided for bringing in high-quality junior and senior facultywith initial research support at a time when few universities could p.o-vide such funding. The Physics Department had 25 faculty members in1965 but, with this grant, the faculty grew to 40 by 1968. The depart-ment has continued to grow and currently has a faculty of 65, including2 Nobel laureates and 5 members of the National Acadcmy of Sciences.

In addition to improving the quality of the faculty, the quality andnumber of the graduate students has also improved. According to pre-sent and former department chairmen, graduate enrollment hasincreased from 100 in 1965 to over 250 in 1985. In addition, postdoc-toral fellows have increased from none in 1965 to Gier 100 in 1985.

Expenditures for the Physics Department have increased from $1.9 mil-lion in 1970 to $10.8 million in 1984. Income from private endowmenthas greatly strengthened the department financially. At the time of ourvisit, the department had six endowed chairs at $1 million each, sixendowed professorships at $100,000 each, and one lectureship. In addi-tion, the University of Texas System has an endowment valued at about$2 billion. Income from the endowment is about $150 million per yearwith two-thirds going to the University of Texas System and one -this dgoing to the Texas A&M System. With this endowment income, the uni-versities pay off bond obligations, finance construction projects, andprovide funds for overall program improvement at the sc'ools.


Chapter 2Roe of Fund4 ng Mechanisms in Improvingthe Quality of University Science

One important feature of the University of Texas at Austin's fundingprocedures is that the university matches federal grant funds desig-nated for equipment. For example, if a researcher in the Physics Depart-ment receives a $100,000 federal grant that includes $20,000 forequipment, the university will provide matching funds for the equip-ment part of the grant. A university official told us this matching proce-dure is a ye y effective method of improving the department's researchprogram.

As mentioned earlier, the Physics Department chairman told us that theNSF Science Development Grant awarded in 1966 was a major factor inthe overall improvement of Texas' Physics Department. However, whenwe discussed with university officials the success of this grant, theycautioned us about the widespread use of this type of funding mecha-nism. School officials told us that the success of development grantsdepends greatly on proper planning for the use of the funds. Forexample, if the funds are used to increase the number of faculty it thedepartment, the university must be able to absorb these faculty costswhenever the grant funds are discontinued. Otherwise, the universitymight have to reduce its faculty and the school would be hack where itwas in the beginning, before the grant funds.

Summary In the development of productive university research organizations,funding mechanisms play different roles at different stages. Thecommon element that was reported to us in improvement at the univer-sities we visited was an explicit commitment from the university toimprove its program and to do so through increases in internal andexternal funding and personnel changes.

Seed funding from either government or private sources was reportedlya prerequisite to program improvement in all of the departments we vis-ited. Two of the five departments we visited received substantial NSFScience Development grants in the late 1960's. University officials atboth schools agreed that the availability of these federal grants was amajor factor in their grogram improvement strategy and enabled eachdepartment to attract excellent researchers, renovate research space,and purchase critical equipment. Although the other three departmentsdid not receive science development grants, they were able to obtainfinancial support from the university, state government, and industry.

After the investment of seed money in the departments we visited,faculty members competed successfully in their fields, and the primary


Chapter 2Role of Funding Mechanisms in Improvingthe Qua lit, of University 3cienee

source of support became the individual project mechanism. Thesemoneys, along with supplemental support from state governor, nt,endowments, industry, or university funds, can generally sustain thequality program, at least in the short run. In the departments we visited,the universities' commitment to absorb the increased faculty costs whenthe science development grant or other seed money ended, helped sus-tain the high-quality programs and allowed the departments time tosecure adequate external funding to make them predominant'y self-sup-porting. The seed money was thus "leveraged" to obtain a broader baseof support.

Page 28 2 5 GAO/RCED-86-75 University Funding

Chapter 3

Role of Funding Mechanisms in thePerformance of Research

The House Science and Technology Committee requested that we assessthe relative merit of different funding mechanisms in terms of theireffects on the productivity and performance of research. While the pre-vious chapter focused on factors affecting the improvement of programquality, this chapter examines the impact of two different fundingmechanisms on the performance of research. We compared five depart-ments that rely prinu.-.-ily on the funding mechanism of individual pro-ject grants with five centers that rely primarily on the fundingmechanism of center support For each department or center, weexamined four key factors that had the potential to affect the perform-ance of researchcoverage of reseal ^h requirements, stability of finan-cial and resource support, the influence of funding mechanisms on theflexibility to pursue new and different categorie, of research, andadministrative burden. (See objectives, scope, and methodology mchapter 1.) While our primary focus was to identify the impact of twofunding mechanisms on these key factors influencing the performanceof research, the case study approach also provided insights into otherinfluences on the performance of research.

We found that particular funding mechanisms, such as individualproject awards, do not by themselves have consistent advantages or dis-advantages for the performance of university research. With few excep-tions, no clear-cut differences emerged between the experience ofcenter- and department-based scientists with federal support. Thenature of the funding and the extent of resource coverage depend uponmany factors, such as differences between agencies, university policies,and varying resource needs. We also found that:

- Distinctions between individual project awards and center funding areblurred by scientists' strategies to increase their ability to performresearch, for example, grant applicat,ons to multiple sources.Certain characteristics of the individual project award mechanism resultin some problems, for example, discontinuous funding for graduatestudents.Issues specific to each field of science, as well as certain characteristicsof funding mechanisms, can impede the performance of research.

The remainder of this chapter highlights findings from our analysis ofthe impact of funding mechanisms and other influences on four key fac-tors with the potential to affect research performance.

Appendix I summarizes the responses of all scientists to selectedquestions.

Page 24 2 r3 GAO/RCED-86-75 Unlvendty Funding

Chapter 3Rale of Funding Mechanisms in thePerformance of Research

Coverage of ResourceRequirements

The performance of research requires continued coverage of resourcerequirements. Scientists need trained technicians, equipment, and spaceto conduct laboratory experiments and other research. Fields of sciencediffer in their resource requirements, depending on the stage of each

eld's devek,ynent and its technological requirements. For example,mathematicians working on "pure" theory may work in isolation withfew assistants and little or no equipment. In contrast, cell biologists toldus they ntay utilize a number of lab assistants, while space scientiststold us they may need large amounts of capital for equipment. In suchlabor- or capital-intensive fields, interruptions or delays in access toresources can slow resea7th progress or force dissolution of establishedresearch teams and laboratories.

We found that while certain funding mechanisms provided more contin-uous access to resources, the design of specific mechanisms seemed tohave less effect on the performance of rft arch than the total volume offunding available for different fields of science and fluctuations in thatfunding. The responses of scientists regarding their ability to acquireneeded resources clustere.' more by fields of science than by experiencewith particular funding mechanisms.

The lack of variation in responses from scientists receiving support fromcenter or individual project awards to cover resource requirementsmight be accounted for by a number of other issues mentioned by thescientists we interviewed. The coverage of resource requirementsreflects interactions between an agency's decisions resulting from itsreview process and policies and an individual scientist's definition ofresource ne .tds for a specific project in P given field of science. Resourcecoverage may be influenced by

the degree of variation among types of support, even within a singlefunding mechanism categoi y;differences in agency review processes;agency policy decisions, such as use of funds to cover equipment orgraduate education;the extent to which universities supplement resources;the types of research undertaken, as well as the scale of researchefforts;individual scientists' perceptions of 'ale extent to which their fundingrequests will be ar proved; andscientists' informal knowledge of what criteria govern decisions madeby agency officials or groups of scientific reviewers.


27

Chapte... 3Role of Funding Mechanisms in thePerformance of Research

These interactions can be better understood in the context of threeresource coverage areas we examined: facilities, equipment, and humanresources.

Facilities and Equipment Experience with individual project or center awards did not appear to bethe significant factor in affecting scientists' responses to questions con-cerning adequacy of equipment and facilities. Instead, perceptions ofproblems in these areas differed by field of science.

Overall, 28 of 36 researchers who had been in the federal award systemsince 1970 said that the quality of facilities for their research hadincreased or 3tayed the same. Scientists in two fieldsplant sciencesand artificial intelligencedid not report decreases in quality of facili-ties since 1970. Scientists reporting decreases were in cell biology, math-ematics, and space science.

Tab la 3.1: FacilitiesFigures in_percentage

Increased Same DecreasedHas the quality of facilities Center 42 9 21 4 35.7 n=14*changed since 1970? Department 54 5 31 8 13.6 n=22

n" here and through the text indicates number of scientists who responded to the question

Differences among fields of science were also seen in equipment cov-erage. Although scientists in all fields, with the exception of mathemati-cians, expressed concern over equipment, space scientists showed themost concern (8 of 11). They told us that much of their equipment is 20years old and is maintained periodically by scientists and technicians. Inaddition, as table 3.2 shows, over half of the scientists stated thatneeded equipment is difficult to obtain. There are no clear-cut differ-ences in the experiences of center and department scientists in the eas.;or difficulty in obtaining equipment.

Table 3.2: EquipmentFigures inpercentage

Agreed DisagreedThe equipment I need is very difficult toobtain under current programs

Cente,Department

54 553 3

45 546 7

n=22n=30

Page 28 28 GAO/W.:ED-8615 liniversity Funding

Human Resources

Chapter 3Role of Funding Mec }umlaut's in thePerformance of Research

The funding mechanisms we looked at were not the most significantfactor influencing responses by scientists to our questions about cov-erage of such human resources as technicians and graduate students.Problems with funding for technicians cut across a number of fields ofsciencecell biology, plant science, artificial intelligence, and space sci-ence. Scientists attributed problems with hiring and retaining techni-cians to factors other than funding mechanisms, such as industrialcompetition and current salary structures for technicians at differentuniversities.

Table 3.3 indicates that both center and department scientists view thisas a problem. Center scientists felt more difficulties with the avaTiabilityof technicians, although both center and department scientists reporteddifficulties in supporting technicians.

Table 5.3: TechniciansFigures in percentage

Increased Same DecreasedHas the availability oftechnicians changed since Center 10.0 30.0 60.0 n=101970? Department 18.2 54 5 27.3 n=11

Agreed DisagreedIt is difficult to support Center 76 2 23.8 n=21technicians needed. Department 86.4 13 6 n=22

Problems cited by scientists relating to funding coverage for graduatestudents touched on a number of interrelated issues concerning univer-sity goals and funding mechanisms available for supporting these goals.We found variations in the types of personnel supported by universityresearch groups. For example, some centers have a clearly definedtraining function, while others support research and not graduate edu-cation. In addition, we found that some problems associated with sup-port for graduate students could be traced to the type of fundingmechanism used. Scientists across all fields (53 of 66) agreed thatproject support should nc., be used to support graduate students as isthe current practice. The negative effects they cited included the disrup-tion caused for graduate students by the loss of support from individualproject awards. They suggested the establishment of separate mecha-nisms for graduate student support.

Page 27 29 GAO/RCED-86-75 Unlveraitr Funding

=11P7....Stability of Financialand Resource Support

Chapter 3Role of Funding Mechanisms in thePerformance of Research

A relatively stable resource and financial environment is generally considered beneficial for the conduct of science. Particularly in resource-intensive areas and ones where teams of researchers must be assembled,the predictability of continued funding is important. The stability ofsupport depends not only on the continuity of funding, but also on itsduration through a project's cycle. To determine the impact of fundingmechanisms and other factors on the stability of support, we 3xamined:the cyclical nature of support, lengthy gaps between periods of funding,and appropriateness of award duration for the research beingperformed.

The Cyclical Nature ofSupport

We found that while center support provided more continuous access toresources, the total volume of funding available for different fields ofscience and fluctuations in that funding seemed to have more of aneffect on the performance of research than the design of specific mecha-nisms. Both center and department z,cientists we surveyed told us theyhave had their federal funding cut (table 3.4). Scientists recognized thecyclical nature of federal support for different topics of research. Scien-tists also recognized the increased opportunities to compete for privatesupport in areas of commercial potential and industry interest, such asartificial intelligence and plant biology in agriculture.

Table 3.4: Funding CutsFigures in Percentage

Yes No

Have you ever had your project Center 77 4 22.6 n=31funding cut? Department 833 16.7 n=36

Scientists in fields of shifting program priorities can also be affected bythe cyclical nature of support. For example, NSF'S attempt to ensure sta-bility at the field of science level in mathematics by dividing availablefunds for the mathematics subfields, such as complex analysis, resultedin destabilizing research environments for certain other subfields andindividuals. This example shows that the effects of funding mechanismson university research cannot be assessed without consideration of con-textual factors such as agency policies.

The influence of factors other than funding mechanisms on the stabilityof the support can be seen in fields of science dependent on mil funding.The Office of Management and Budget proposed cutting the number ofNIH awards from 6,529 in fiscal year 1985 to 5,000 new and continuingawards in fiscal year 1986 and further to use the savings from that

Page 28



reduction to spread the availthle funds by distributing the awards over2 or more years instead of 1 year. Scientists in cell biology, one of thefields supported by NIH, told us they were concerned with the pollticiza-don of federal funding for research (e.g., we heard corm' _nts such as"non-scientific events at the federal level," "arbitrary OMB decisions,"and that fluctuations "depelid on the Administration"). Their percep-tions of instability are indicated by the contrast between their success inobtaining funding and an increased sense of unprediLability (table 3.5).

Table 3.5: Changes Over the Last 15Years In Areas Affecting ResearchPerformance

Figures in PercentageIncreased Same Decreased

How has the predictability orobtaining federal projectfunding changed')

How has your success rate infunding changed?

Center 23.1 15.4 61.5 n=13Department 27 3 27.3 45 5 n=22

Center 10.0 80 0 10 0 n=10Department 11.1 66 7 22 2 n=18

Funding Gaps We found that the type of funding mechanism used had a more signifi-cant impact in the area of funding gaps than in other areas related tostability. For departmental scientists who received individual projectawards, rather than center funding, funding gaps sometimes translatedinto ending support that broke up research teams and caused the loss oftrained professional technicians. Scientists noted that the social and eco-nomic costs of funding gaps (human suffering, retooling, increased timeexpended by scientists in the day-to-day operations of the lab) were anintangible cost in the performance of research.

In contrast, we found that the center mechanism provided a measure offlexibility that enhanced the stability of the research environment forthose scientist; who received center support. Scientists cited theinformal sharing of resources possible under center funding as one con-tributing factor to stability of funding. Center funding provides someseed money to start research that would otherwise oe unfunded andbridges periods when noncenter funds are terminated. Finally, it canprovide for more continuous support of professional technicians.Funding gaps in the centers were seen as delays in funding, rather thanrs an end to support.

Although center support provided more stability in funding, we foundthat some department scientists had developed strategie? that seemed tocompensate for funding gaps. To prevent an abrupt stop to their

Page 2931

GAO/RCRD-86-76 University Funding


research, scientists will apply to multiple sponsors in order to guaranteethe continuity of their work. When one project ei.ds, the researcher isstill receiving support from other sources. A second device is the prac-tice of working as a co-investigator on Imeone else's award. To meetequipment needs, scientists in one department we visited collaboratedand were able to pool resources from various project awards in order toestablish eciipment for common use.

Table 3.6: Funding GapsFigures in

PercentageYes No

Have funding gaps been a problem? Center 27.6 72 4 n=29Department 50 0 50 0 n=34

Award Duration Scientists receiving both types of mechanisms expressed concern aboutaward duration (table 3.7). However, scientists in most of the centers westudied commented that they had a longer term commitment under thecenter mechanism than scientists who received individual Projectawards. Award duration affects stability because award periods do notalways match the actual time needed to perform research, which canvary even within a field. For example, one scientist told us that bio-chemistry projects take considerably less time to complete than geneticmanipulation experiments in agriculture, where scientists must allow acomplete regeneration of crops before testing can take place. Scientistsalso suggested that for many fields, shorter duration awards (less than 2years) did not recognize start-up time as a legitimate facet of researchand thus did not permit the following of coherent research strategies.Finally, scientists recognized the difference between the long-term wayin which they perceive research (scientists conceptualized their work aslife long, or in terms like "a 50-year project") and the relatively short-term way in which funding agencies perceive research (in 3-to 5-yearincrements).

Table 3.7: Experience With FederalAwards

Award periods are too short to finish aproject within one award cycle

There's not enough time to completescholarly articles during the projectaward period.

Page 30

Figures in PercentageAgreed Disagreed

Center 59 3 40 7 n=27Department 61 8 38 2 n=34

Center 45 2 54 8 n=31Department 545 45 5 n=33

32GAO /RCED -8&75 University Funding

Chapter 3Role of Funding Mechanisms in the'Performance of Research

Types of Research Some differences in the types of research supported emerged betweenthe two mechanisms studied. One criticism of the individual projeaaward review system is that it does not adequately support innovative,high-risk research. A task force of the National Science FoundationAdvisory Council identified the following three classes of inno ative,high-risk proposals: research that challenges currently accepted scien-tific hypotheses; interdisciplinary proposals or research that transfersknowledge from one scientific field to another; and research that is atthe edge of technical feasibility. To determine which mechanisms (cen-ters or individual project awards) more often support innovative, high-risk, and interdisciplinary research, we asked scientists a series of ques-tions about their research.

We found that more scientists in centers are likely to perform the typesof research defined as innovative, high risk, or interdisciplinary. Morecenter than departmental scientists:

performed research bridging two or more fields (30 of 32 center scien-tists versus 21 of 36 departmental scientists);proposed research into new areas (25 of 32 center scientists versus 14 of33 departmental scientists); andproposed work with industrial applications (9 of 32 center scientistsversus 3 of 33 departmental scientists).

Although innovath e, high-risk, and interdisciplinary research tended tobe performed by scientists in centers, in certain cases the field of sci-ence, not the affiliation with a center or department, seemed to influ-ence the types of research performed. For example, all plant scientists inthe center and department (11 of 11) described their research as inter-disciplinary, bridging two or more fields. Differences were not clear cutbetween scientists who proposed new technical processes with supportfrom the center or individual pros'oct awards. Few mathematicians hadproposed new technical processes (3 of 20) or proposed research intonew areas (7 of 13). In contrast, almost all plant scientists (10 of 11) andscientists in artificial intelligence (8 of 9) had proposed research in newareas.

Administrative Burden One aspect of the current reliance on the individual project awardsystem that has been criticized by scientists is the time and expense ofpreparing and administering a large volume of applications. Time spentby scientists in preparing and reviewing research proposals is seen as

Fage 31 GAO/RCED436-75 University Funding

33


resulting in a decline of research productivity. Discussion has also sug-gested a need to streamline procedures for administering grants andcontracts, without reference to the particular funding mechanisminvolved.

The time commitment by scientists required to participate in the federalfunding system cart be divided into two categories: preaward and post-award. This time encompasses not only proposal applications, but alsoresponses to sponsoring agencies' requests for proposal review, partici-pation in technical monitoring, and the preparation of status and finalwork reports.

We examined the relative amount of time spent in award-related activi-ties by scientists receiving center support and those departmental scien-tists receiving support from individual project awards. We were alsointerested in whether scientists perceived differences in administrativeburden between sponsors. We also asked university administrators tocomment on these issues.

We found that, for the scientists we interviewed, the amount of timespent applying for awards, responding to award requirements, andreviewing proposals varied not by type of mechanism but more by thefield of science and the requirements of the dominant agency sponsoringresearch m each field. We also found that no single issue emerged amongthese 70 scientists regarding the presence of administrative burdenScientists' perceptions of difficulties in this area can be shaped by anumber of factors: whether individuals or groups submit multiple appli-cations in order to obtain federal a- yards, the number of researchers inrelationship to available funding, and changes in agency requirements.We found that scientists at the schools we visited tended to cite anumber of problems when specifically asked about administrativeburden, ranging from the time spent in responding to regulitionsimposed by different governmental bodies to time a.'.d effort reporting.

Page 32 3q GAO/RCM-8645 University Funding


Table 3.8: Average Time Spent byScientists In Award-Related Activities Writing Proposal

applications review days/weeks/year year

Statusreports

days/year

Technical Noncompetitivemonitoring renewal days/days/year year

FIELD OFSCIENCE

Plant science 5 6 18 5 6 3 3 6 3 7n=10 n=11 n=11 n=10 n=9

Cell biology 4 7 15 5 5 4 31 1 1

n=12 n=10 n=14 n=14 n=14Aathematics 2 0 5.9 2 6 8 1 9

n=20 n=20 n=18 n=16 n=14Space science 3 5 7 7 4 1 1.2 3 5

n=11 n=12 n=11 n=11 n=11

Artificial 3 9 9 7 6 6 6 4 1 9intelligence n=10 n=10 n=9 n=9 n=9All scientists 3 6 10 6 4 i 2 7 2.3

n=63 n=63 n=63 n=60 n=57IIIIIKi

Table 3.8 lists differences among fields for the 10 schools we visited inthe amount of time spent in activities. Differences result from variationin agency requirements for funding research rather than from the typeof mechanism employed. The major distinctions among fields seemed tobe in the area of preaward and postaward requirements. Scientistsreceiving funding from the Department of Defense, the National kero-nautics and Space Administration, and the Department of Energy (agen-cies that make decisions internally or through combined internal andexternal review) might srend less time on proposal review, one exampleof a preaward requirement, than scientists supported by NSF and NIH. NSF

and NIH use only one form of decision making, peer review, a processdesigned to have groups of scientists to review the merits of work pro-posed by colleagues in various specialties. In contrast, researchers inartificial intelligence spent more time responding to the requirements oftechnical monitors, a postaward requirement common m researchfunded 'ay the Department of Defense. Three scientists in artificial intel-ligence, a field that receives support from both civilian and defenseagencies, perceived NsF to be the most burdensome in preaward require-ments and least demanding in postaward requirements compared todefense 2.6encies.

While there were no clear-cut differences overall in the administrativerequirements, we found that SOT' . centers are designed in such a way asto insulate staff from the burden of administrative tasks. For example,

Page 88 35 GAO/RCED-8676 University Funding


"sble 3.9: Shifting Funds

at one university the center director had a small core staff to handle thewriting of proposals and other award-related tasks.

One postaward issue we spe"ifically addrecsed concerned the ease ordifficulty in shifting funds between expenditum categories (table 3.9).We asked researchers whether they found it difficult to shift fundsbetween categories. We wanted to know whether they had the flexibilityto shift resources in the event of unexpected events such as a change inthe direction of their research. This did not seem to be a clear-cut issuefor center investigators, who split on their responses to this question. Incontrast, more department scientists (25 of 34) found it easier to shiftfunds. Certain restrictions seem to lead some researchers to resort toother sources of funding rather than attempt to acquire approval forsuch expenses as travel or equipment. However, several researcherspraised NIH and NSF, agencies that have decentralized administrativeresponsibility for overseeing shifts in expenditures to the universitylevel. We also found examples of unique forms of the individual projectaward that are flexible in character, such as general research contractsfrom the National Aeronautics and Space Administration and the Officeof Naval Research.

General research contracts have broad objectives and provide the prin-cipal investigator with considerable discretion in how the funds areused. Among other uses of these contracts, the principal investigator cansuppert young investigators who have not established a performancerecord or technicians and graduate students during funding gaps.

Figures in Percentage_Agreed Disagreed

It is difficult to shift funds Center 50 0 50 0 n=22between expenditure categories Department 26 5 73 5 n=34

For university administrators, three factors affect the amount of timespent in administering federal research awards. Administrative time canbe increased by institutional policies for review, differences in the pro-cess of negotiating and administering contracts with different sponsors,and difficulties with specific legal instruments rather than fundingmechanisms.

Summary Our case studies of the role of different funding mechanisms inenhancing or inhibiting research performance show that particular


36


funding mechanisms we looked at do not always have consistent advan-tages or disadvantages in the performance of research. Performance ofresearch can be affected by any of the following factors: resource cov-erage, stability, the flexibility to pursue new research ideas, and admir-istrative burden. For these factors, we found issues that were eitherfunding mechanism-related, field of science-related, or cut acrossfunding mechanisms and fields of science.

In looking at the issues that relate to specific funding mechanisms, thecenter grants we examined were somewhat more likely to provide morecontinuous access to resources; to afford a greater degree of stability forthe performance of research; and to enhance the performance of innova-tive, hie- -risk, or interdisciplinary research.

Field of science-related L ..ies included he following: the cyclical natureof support for the field, changes in agei, relationships, and the uniqueneeds of subfields. The cyclical nature o support for different fieldsseemed to explain differences in rrsource coverage between fields. Dif-ferences among fields of science were seen in coverage of resourcesfacilities and equipment. For example, scientists in organizationsreceiving a relatively rapid increase in volume of funding, such as artifi-cial intelligence and plant biology, said that the quality of facilities fortheir research had increased or stayed the same. Space scientists,working in a field with stable or decreasing funding, showed more con-cern over the condition of their facilities and equipment. Cell biology is afield of science that illustrates the effects of a change in agency relation-ships. In this field, which is primarily supported by NIH, scientists weinterviewed described the destabilizing of their research environmentcaused by executive branch decisions to change the number of awardsmade by NIH for individual project support. The unique needs of sub-fields can also affect scientists' experience with funding mechanisms.For example, the time needed to perform research can vary even withina field as in the case of plant biology in which it may take several yearsfor a new crop to grow and be tested.

Issues that cut across mechanisms and fields of science include the cur-rent problem of finding and keeping technicians. Similarly, perceptionsof administrative burden seemed influenced by factors other than mech-anisms and characteristics of a field of science. Problems were attrib-uted to a range of factors, including university procurement policies andstate and municipal regulations.

Page 35 37 GAO/RCED46-75 University Pending

Appendix I

Summary of All Scientists' Responses toSelected Questions

Figures in Percentage

Stability of Financial and Resource SupportHas the success rate in funding of federal proposals over the last 15 years changed?(n=28)Increased 17 9Same 71 4Decreased 10 7

Award periods are too short to finish a project within one award cycle. (n=61)AgreedDisagreed

60.739 3

There is not enough time to complete scholarly articles during the project award period(n=64)AgreedDisagreed

50.050.0

Have you had problems because of gaps in your funding? (n=63)YesNo

39.7603

Has the predictability of obtaining federal project funding changed over the last 15 years?(n=35)IncreasedSameDecreased

25.722.951 4

Have you ever had your project funding cut? (n=67)YesNo

80.6194

Coverage of Resource RequirementsHas the qual,ty of facilities changed since 19702 (n=36)Increased 22 2Same 27.8Decreased 50 0

The equipment I need is very difficult to obtain under current federal award program (n=52)Agreed 53.8Disagreed 46 1

Has the availability of technicians changed since 1970? (n=21)increased 14.3Same 42.9Decreased 42 9

It is difficult to support technicians needed (r,=43)Agreed 81.4Disagreed 18 6

Types of ResearchSome projects are no funded because they don't fit conventional areas hvored byreviewers. (n= 55)Agreed 41.8Disagreed 58.2

Administrative BurderiIt is difficult to shift funds between expenditure categories (n=56)Agreed 35.7Disagreed 64 3

"n" indicates the number of scientists who responded to the question

(006710) Page 36 38 GAO/RCED-86-75 University Funding

Requests for copies of GAO reports shoald be sent to:

U.S. General Accounting OfficePost Office Box 6015Gaithersburg, Maryland 20877

Telephone 202-275-6241

The first five copies of each report are free. Additional copies are$2.00 each.

There is a 25% discount on orders for 100 or more copies mailed to asingle address.

Orders must be prepaid by cash or by check or money order made out tothe Superintendent of Documents.