United States General Accounting Office Washington, DC. 20648 Human Resources Division B-236667 February 21,199O The Honorable Tom Harkin Chairman, Subcommittee on Labor, Health and Human Services, and Education Committee on Appropriations United States Senate Dear Mr. Chairman: This report responds to your August 11, 1989, request for information on the Pell grant program. This program is administered by the Depart- ment of Education and provides grants to financially needy students pursuing a postsecondary education. You requested that we describe the Department’s process for estimating Pell grant program costs, and determine what factors caused or contrib- uted to the Department revising its fiscal year 1990 budget request. You were particularly concerned about the Department’s latest estimate, which was $696 million higher than its initial request for fiscal year 1990. In addition, you asked us to determine whether the Department has authority to borrow funds from subsequent years’ appropriations to pay for Pell grant funding shortfalls from previous years. On January 16,1990, we briefed your staff on the results of our analy- sis. We reported that underestimating the number of recipients was the primary cause of the increase in the Department’s request. This briefing report summarizes the information presented at that meeting (see am. I>- Pell grants are available to students meeting certain eligibility criteria, with the amount of aid determined by their financial need, educational costs, and whether they attend school full or part time. Students are eligible for grants for up to 6 years of study, and they must reapply each year. The number and amount of Pell grants available to students are limited by the total funds the Congress annually appropriates and the number of applicants who are eligible to receive the grants. The Pell grant program is funded so that appropriations for one year fund the program for the subsequent year. For example, the fiscal year 1990 appropriations will be used to disburse PeIl grants to students Page 1 GAO/HI&D90-73BR Estimating Pell Grant Program Costs B-236667 attending school during the 1990-91 award year. An award year covers the period beginning July 1 and ending on the subsequent June 30. Since the program’s inception in 1972, the number of recipients has increased from 176,000 to 3.2 million in fiscal year 1988, while funding has increased from $122 million during that first year to $4.3 billion in 1988. In award year 1988-89, the program awarded grants ranging from $200 to $2,200, with an average award of $1,397. To determine how the Department estimates Pell grant costs and what Scope and factors affect these costs we (1) interviewed key personnel responsible Methodology for developing the Department’s Pell grant budget estimates and (2) determined what methods and factors the Department used to develop the estimates. We analyzed the Department’s budget requests and asked Department officials to explain why they estimated additional costs of $696 million for fiscal year 1990. We interviewed officials of the Congressional Budget Office and the American Council on Education (ACE), organizations that also periodi- cally estimate Pell grant program costs, to obtain their views on the Department’s budget formulation process. We also interviewed Office of Management and Budget officials who provide economic indicators, such as family income and the consumer price index, to federal agencies for use in developing their budget requests and review the Department’s budget estimates on behalf of the administration. Further, we asked the Department for its opinion on what legal basis or legislative authority it has to borrow against future years’ appropria- tions. We evaluated their response and issued our report to you on Janu- ary 26, 1990. The Department uses a three-step process to develop its Pell grant How the Department budget estimate: a computer cost model, a series of adjustments to of Education Develops reflect program changes that the model does not include, and budgetary Its Budget Estimates adjustments. For the 1990 budget request, the model used over 30 eco- nomic indicators. It also used data such as income, dependency, and cost of education, from a sample of about 151,000 students who applied for and may have received Pell grants during the 1986-87 award year. This was the latest program data available at the time the Department built its estimate. Page 2 GAO/HRD-90-73BR Estimating Pell Grant Program Costa B-236667 After the Department uses the model to develop the first-step estimate, it manually adjusts the model’s estimate to reflect recent program changes that the model is unable to simulate. For example, in developing its fiscal year 1990 budget request, the Department made three program adjustments to the model’s estimate to account for program changes made by the Higher Education Amendments of 1986 (Public Law 99- 49&), which (1) expands the definition of an independent student (gener- ally a person responsible for his or her own financial support), (2) pro- vides school financial aid administrators’ discretion in adjusting a student’s cost of attending school and the amount of funds a student’s family could be expected to contribute to his or her education, and (3) makes students attending school less than half-time eligible for Pell grants. The Department uses a third step to adjust its estimate for possible pro- gram funding shortages and surpluses+ The resulting amount becomes the Department’s request to the Congress. For example, the initial 1990 budget request included an additional $96 million to adjust for the esti- mated difference between funding and program courtsfor fiscal years’ 1988 and 1989. This step also includes adjustments made to reflect esti- mated savings from the Department’s legislative proposals. (Appendix II describes the Department’s process for estimating Pcll grant program costs in more detail.) Cost Model Estimate Historically, the Department’s cost model estimate has been a reasona- Closely Predicted Actual bly accurate predictor of program costs, From award years 1980-81 through 1988-89, the model’s estimates were within an average of 3.63 Program Costs percent of actual costs. Department officials said that if the award year 1983-84 were not included in this analysis-when there was a slight recession which they could not predict-the average difference between the cost model estimate and actual costs would have only been 2.25 per- cent. (See fig. 1.) However, in each of those years the Department reduced the model’s estimate by making program and budgetary ad.justments. These adjust-, ments generally resulted in a budget request being submitted to the Con- gress that underestimated that year’s actual costs. Department officials said that a major reason for reducing the cost model estimate during these years was to reflect estimated program savings. The Department expected to achieve t.hese savings from legislative proposals it submit- ted as part of its annual budget requests. Page 3 GAO/HRD-90-73BR Estimating Pell Grant Program Costs B-236667 Figure 1: Final Budget Requests Generally Less Than Actual Program 6.0 Dollars In Billions costs 5.6 5.2 4.6 4.4 4*0 3.6 32 2.6 2.4 2.0 1.6 1.2 0.8 0.4 0 Award Year - Cost model estimate I - - - Actual program cost m Final budget request Appendix III lists the dollar values shown in the above chart. The Department’s second budget request for fiscal year 1990 (submitted What Caused the $696 in June 1989) was actually $522 million more than its first request (sub- Million Increase? mitted in January 1989). The Department estimated it would need an additional $696 million to operate the program under existing legisla- tion. However, it requested $174 million less than this estimate in antici- pation of savings expected to accrue from legislative changes the Department was proposing. The $522 million net increase was attributed to (1) an increased cost model estimate of $560.7 million based on revised economic indicators and an unexpected increase of 500,000 grant recipients, (2) an addi- tional budgetary shortage of $60.7 million and (3) a decrease of $99.4 million in program adjustments, largely due to a lower estimated cost for Page 4 GAO/HRD90-73BR Estimating Pell Grant Program Costs B236667 financial aid administrators’ discretion. Table 1 compares the Depart- ment’s initial and most recent fiscal year 1990 Pell grant budget requests. Table 1: Comparison of the Department’s Fiscal Year 1990 Pell Grant Budget Dollars In millions ._,...-. Requests January 1989 June 1989 request request Change Cost model $4,274.5 $4,835.2 +$560 7 Program adjustments 369.7 270.3 .- -99.4 Budget adjustments 95.8 156.5 + 60.7 Final budaet reauest $4,740.0 $5,262.0 +$522.d Appendix IV contains more information on the comparison of the Department’s fiscal year 1990 budget estimates. The Department attributed the increased number of students who were expected to receive Pell grants to (1) amendments to the Higher Educa- tion Act requiring all students to apply for a Pell grant, whether or not they are eligible before they can receive other federal aid, such as a Stafford student loan, and (2) an increase in the number of students applying for and receiving Pell grants. The Department also estimated that independent students--who generally receive larger grants than dependent students-increased by 9 percent from the preceding award year, and many of them will probably qualify for Pell grants. Page 6 GAO/HRD9@73BR Estimating Pell Grant Program Costs - B-236667 As agreed with your office, we did not obtain written comments from the Department of Education. We did, however, discuss its contents with Department program officials and incorporated their comments where appropriate. We are sending copies of this report to other congressional committees, the Department of Education, the Office of Management and Budget, the Congressional Budget Office, the American Council on Education, and other interested parties. If you have questions about the information presented, please call me on (202) 275-1793. Other major contributors to this report are listed in appendix V. Sincerely yours, Franklin Frazier Director, Education and Employment Issues Page 6 GAO/HRD-90-73BR Estimating PelI Grant Program Costs Page 7 GAO/HRD-90-73BR Estimating Fell Grant Program Costs Contents Letter 1 Appendix I 12 Pell Grants: How the Department of Education Estimates Program Costs Appendix II 33 Department of How the Computer Model Works 33 Program Adjustments 34 Education’s Process Budgetary Adjustments 35 for Estimating Pell Grant Program Costs Appendix III 36 Table Supporting Figures in Report Text Appendix IV 37 Department of January 1989 Estimate 37 Education’s Fiscal February 1989 Estimate 38 April 1989 Estimate 38 Year 1990 Pell Grant June 1989 Estimate 38 Program Budget Comparison of January and June Budget Requests 39 Request Appendix V 41 Major Contributors to This Briefing Report Related GAO Products 44 Tables Table 1: Comparison of the Department’s Fiscal Year 5 1990 Pell Grant Budget Requests Page 8 GAO/HRD-SW73BR Estimating Pelt Grant Program Costs Contents Table III. 1: Comparison of Pell Grant Cost Model 36 Estimates, Actual Program Costs, and Final Budget Requests: Award Years 1980-81 to 1988-89 Table IV.1: Comparison of Education’s Fiscal Year 1990 37 Pell Grant Budget Estimates Table IV.2: Comparison of Award Year 1990 Pell Grant 39 Applicants and Recipients in the Department’s January and June 1989 Budget Requests Table IV.3: Major Cost Components for the January and 40 June 1989 Pell Grant Budget Requests Figures Figure 1: Final Budget Requests Generally Less Than 4 Actual Program Costs Figure I. 1: Pell Grants 12 Figure I. 2: Background 13 Figure 1.3: Timeframes for FY 1990 Pell Grant Program 14 Budget Request Figure 1.4: Objectives 15 Figure 1.5: Scope 16 Figure 1.6: Methodology 17 Figure 1.7: What Did We Find? 18 Figure 1.8: Department Uses Three-Step Process for 19 Budget Request Figure 1.9: How Does the Cost Model Work (FY 1990)? 20 Figure 1.10: Model Estimates, Budget Requests, and 21 Actual Costs Figure I. 11: Program Adjustments in FY 1990 Budget 22 Estimate Figure 1.12: Budgetary Adjustments in FY 1990 Budget 23 Estimate Figure 1.13: Major Milestones in FY 1990 Pell Grant 24 Budget Estimates Figure 1.14: FY 1990 Budget Process Was Atypical 25 Figure 1.15: What Happened to FY 1990 Budget Estimates? 26 Figure 1.16: January 1989 Estimate 27 Figure 1.17: February 1989 Estimate 28 Figure 1.18: April 1989 Estimate 29 Figure I. 19: June 1989 Estimate 30 Figure 1.20: Last Request Was Actually $522 Million 31 Higher Than First Figure 1.21: What Caused the $522 Million Budget 32 Estimate Increase? Page B OAO/HRIMtO~73BR Estimating Pell Grant Program Costs Contents Abbreviations ACE American Council on Education CBO Congressional Budget Office FY Fiscal year GAO General Accounting Office OMB Office of Management and Budget SrS Supplemental Loans for Students Page 10 GAO/IiRDfW73BR Estimating Pell Grant hogram Costs Page 11 GAO/HRD-!%I-73BR Estimating PeU Grant Program Costs Appendix I Pell Grants: How the Department of Education Estimates Program Costs Figure 1.1 G&) Pell Grants How the Department of Education Estimates Program Costs L Page 12 GAO/HRD90-73BR Estimating Pell Grant Program Costs Appendix I Fell Grants: How the Department of Education Estimates Program Costs Figure 1.2 GJU Background 0Grants go to most needy l Maximum grant for 1988-89 award year was $2,200 l 3.2 million recipients aAbout $4.3 billion funded l Budget estimate for fiscal year (FY) 1990 was revised three times and last estimate was $696 million higher than first Page 13 GAO/HRD-90-73BR Estimating Pell Grant Program Costa Appendix I Pell Grants: How the Department of Education Estimates Program C!!sts Figure 1.3 GAO Timeframes for FY 1990 Pell Grant Program Budget Request l Fiscal and award years differ lFiscal year covers period Oct. 1989 to Sept. 1990 mAward year covers period July 1990 to June 1991 l FY 1990 budget request 4Jsed 1986-87 award year data (finalized in Dec. 1987) l Prepared in summer of 1988 @Projected Pell grant costs for 1990-91 award year Page 14 GAO/HIRD-90-73BR Estimating Pell Grant Program Costs Appendix I Pell Grants: Bow the Department of Education Estimates Program Costa Figure 1.4 GAQ Objectives GAO was asked to determine l How the Department estimates Pell grant costs, including what factors affect costs? l What caused the $696 million increase? l Whether the Department can borrow against future appropriations? Page 16 GAO/IiRDW73BR Esthnating Pell Grant Program Cbstu Appendix I PeLl Grants: How the Department of Education Estimates Program Costs Figure 1.5 GAQ Scope 9 Determined the Department’s budget process for Pell grants l Determined factors used in making budget estimates l Reviewed components of budget estimates mReviewed the Department’s authority for future borrowing Page 16 GAO,‘I3ltD~7BBR Estimating Pell Grant Program Coutr Appendix I PelI Grants: How the Department of Education EBtimates Program Costs Figure 1.6 G&I Methodology l Compared estimated costs, budget requests, and actual program costs l Interviewed Department, CBO, OMB, and ACE officials l Requested legal opinion from the Department on future borrowing l Past authority l Current authority Page 17 GAO/HEUMO-‘73BR Estimating PeB Grant Program Costs Appendix I Pell Grantrr: How the Lkpartment of Education Estimates Program Costa Figure 1.7 0 What Did We Find? l Department uses a three-step process to estimate costs *Computer cost model l Program adjustments aBudgetary adjustments l Underestimating recipients primarily caused increase . GAO responded separately on the Department’s borrowing against future appropriations J Page 18 GAO/HRD9@73BR Estimating Pell Grant Program Costs Appendix I PeU Grants: How the Department of Education Estimates hgram Costa Figure 1.8 w Department Uses Three-Step Process for Budget Request l Uses a computer cost model to prepare initial estimate l Adjusts the model’s results to reflect recent legislative (program) changes that the model is unable to simulate l Adjusts the model’s results for expected budgetary shortages or surpluses, resulting in a final budget request Page 19 GAO/~W73BE Estimating PeU Grant Program Costs Appendix I Pell Grants: How the Department of Education Ektimates Program Casts Figure 1.9 GAQ How Does the Cost Model Work (FY 1990)? l Contained data on sample of 151,000 applicants/recipients from award year 1986-87 l Used over 30 economic indicators, such as income and the consumer price index Page 20 GAO,JHRIMM~BR Estimating PeU Grant Program Costs Appendix I Pell Granti How the Department of Education Estimates Program Costs Figure 1.10 w Model Estimates, Budget Requests, and Actual Costs For award years 1980-81 to 1988-89 l Model estimates exceeded actual costs by an average difference of 3.63 percent l Budget requests reduced from model estimates l Actual costs greater than budget requests - Page 21 GAO/IiIUWO-73BR Estimating Pell Grant Program Costs Appendix 1 PeU Grants How the Department of Education Estimates Program &stS Figure 1.11 w Program Adjustments in FY 1990 Budget Estimate l Revised definition of independent student l Allowed financial aid administrator discretion l Awarded grants to less than half-time students Page 22 GAO/HRD90-73BR Estimating PeU Grunt Program Costs Appendix I Pell Grants: How the Department of Education Estimates Program CuMa Figure 1.12 GAQ Budgetary Adjustments in FY 1990 Budget Estimate l Budgetary shortage for the January, February, and June estimates .$95.8 million, $3 million, and $156.5 million, respectively l Budgetary surplus of $55 million for the April estimate Page 23 GAO/HBD-90.73BR Estimating Pell Grant Program Costs Appendix I Pell Grants: How the Department of Education Estimates Program Costs Figure 1.13 GA) Major Milestones in FY 1990 Pell Grant Budget Estimates Five key dates l June 1987: end of award year 1986-87 l Aug. 1988: prepare budget using 1986-87 data l Jan. 1989: President’s budget to Congress l June 1989: OMB mid-session review l Jan. 1990: final adjustments to FY 1990 budget Page 24 GAO/HRINW73BR Estimating Pell Grant Program Costs Appendix I PeB Grants: How the Department of Education Estimates Ehgram Casts Figure 1.14 GAQ FY 1990 Budget Process Was Atypical l Budget estimate normally submitted to Congress once a year in January with no update l Four budget estimates were prepared for FY 1990 (January, February, April, and June 1989) Page 25 GAO,MRD90-73BE Estimating Pell Grant Program Costs Appendix I Pell Grants: How the Department of Education EMmates Program Costa Fiaure 1.15 GAD What Happened to FY 1990 Budget Estimates? Estimates kept changing l January 1989: $4.740 billion l February 1989: $4.592 billion l April 1989: $4.517 billion l June 1989: $5.262 billion Page 26 GAO/HRINO-73BR Estimating PeU Grant Progmn Costs Appendix I PeB Grants: How the Department of Education Estimates Program Costs Figure 1.16 GAQ January 1989 Estimate l Expected recipients were 3,142,OOO l Based on June 1988 economic indicators (November 1988 indicators were available) l Model estimate was $4.274 billion and budget estimate was $4.740 billion Page 27 GAO/HRD9@73BR Estimating Pell Grant Program Chats Appendix I Pell Grants: How the Department of Education Estimates Program Costs Figure 1.17 GM February 1989 Estimate l Expected recipients were 3,105,334 l Based on November 1988 economic indicators l Model estimate was $4.222 billion and budget estimate was $4.592 billion i Page 28 GAO/HRD-90-73BR Estimatbg PeU Grant Program Costs AppendixI Pell Grants: How the Department of Education Estimates Program Costa Figure 1.18 0 April 1989 Estimate l Expected recipients were 3,261 ,134 l Based on November 1988 economic indicators l Model estimate was $4.205 billion and budget estimate was $4.517 billion Page29 GAO/HRD90-73BR EstimatingPeUGrantProgramCosti Appendix I Pell Grants: How the Department of Education Estimates Program Costs GAQ June 1989 Estimate l Expected recipients were 3,640,895 l Based on June 1989 economic indicators l Model estimate was $4.835 billion and budget estimate was $5.262 billion Page 30 GAO,‘IUDW73BR Estlmatlng Pell Grant Program caste Appendix I PeB Grants: How the Department of Education Estimates Program Costs Figure 1.20 m Last Request Was Actually $522 Million Higher Than First l June 1989 estimate was $5.436 billion l This estimate was reduced to reflect proposed legislative savings of $174 million 0 Net request was $5.262 billion (compared with January estimate of $4.740 billion l Difference is $522 million i Page 31 GAO/HRD-90-73BR Estimating PeU Grant Program Costs Appendix I Pell Grants: How the Department of Education Estimates Program Ckxxts Figure 1.21 G&I What Caused the $522 Million Budget Estimate Increase? 500,000 more recipients l More grant recipients @Students must apply for Pell grant before Stafford or SLS l More independent students l 9% growth from FY 1988 aTwice the increase from 1987 to 1988 Page 32 GAO/HRB90-73BR Estimating Pell Grant Program Costs Appendix II Department of Education’s Processfor Estimating Pell Grant Program Costs The Department of Education performs a three-step process to prepare Pell grant program budget estimates. It uses a computer cost model, which incorporates a number of economic and population indicators, to produce a preliminary estimate. The Department then manually adjusts the model estimate to reflect the impact of recent program changes, which the model cannot simulate. It further adjusts the resulting esti- mate for budgetary shortages or surpluses-to reflect differences in funding and program costs from previous estimates-to determine its final budget request. The Pell grant cost model is a computer simulation procedure for esti- How the Computer mating the number of Pell grant recipients and program costs. The Basic Model Works Grants Branch in the Department’s Division of Policy and Program Development created the model, annually updates it, and has used it to estimate program costs since the program’s inception in 1972. The Department selects a random sample of Pell grant applicants and recipients from a combined file to estimate the number of applicants eli- gible for a Pell grant, whether they are dependent on others for their support or provide their own, and the amount of grants awarded for the budget year. For example, in preparing the fiscal year 1990 budget request the Department used a random sample of 151,000 applicants/ recipients from award year 1986-87, which began on July 1, 1986, and ended on June 30, 1987. The 1986-87 award year was the latest for which the Department had complete data when it began preparing its fiscal year 1990 budget during the summer of 1988. Each sampled record is weighted (adjusted) with various economic indicators provided by the Office of Management and Budget for esti- mating changes in prospective students’ income and assets, resulting in a student aid index, The index measures eligibility and calculates Pell grant awards and expected disbursements for eligible students. The probability of these students receiving awards is factored into the calcu- lation, resulting in estimated dollar amounts and number of recipients, Three organizations use the model’s database and sample of applicants and recipients but each organization uses them for different purposes. The Department’s Office of Planning, Budget, and Evaluation reviews changes the Division of Policy and Program Development makes to the model to reflect expected funding shortages or surpluses. The American Council on Education uses the model to project costs of program changes impacting on the schools it represents, while the Congressional Budget Office uses the model to develop program cost estimates by applying its Page 33 GAO/HRD-99-73BR Estimating Pell Grant Program Costs Appendix II Department of Education’s Process for Estimating Pell Grant Program Costs own economic indicators to the Department’s sample of applicants/ recipients. Limitations of the The Department’s Pell grant program computer cost-modeling process has several known limitations that may affect its ability to accurately Computer Modeling predict program costs, although in practice it has been reasonably accu- Process rate. These limitations are a result of the timing of various milestones in the program’s budget cycle. The Department begins to prepare its budget request 2 years in advance of the award year using the latest available information which is a year old. In part, this 2-year advance preparation results from a requirement that the Department submit its budget request for a fiscal year period that begins 9 months before the start of the program’s award year. In addition, the Department adjusts its model estimate to recognize the effect of program changes for which little data are available. The Department’s model uses data that are a year old because complete data for the most recent award year are not available when it begins to prepare fiscal year budgets. For example, when the Department pre- pared its fiscal year 1990 budget in the summer of 1988, award year 1986-87 was the latest complete year for which data were available. Complete data from award year 1987-88 were not available because the Department does not close out its Pell grant recipient file until 6 months following the end of the award year. Despite these limitations, the cost model estimate has been a reasonably good predictor of Pell grant costs, varying from actual costs by an average of only 3.63 percent for award years 1980-81 to 1988-89. s The Department made manual program adjustments to the cost model Program Adjustments estimate for fiscal year 1990 to recognize legislative changes to the pro- gram that were beyond the model’s capability to simulate, since little historical data were available. Adjustments made accounted for legisla- tive amendments that (1) changed the definition of independent student, (2) modified financial aid administrators’ discretion to determine stu- E dents’ cost to attend school and their expected family contribution, and (3) expanded the eligibility of students enrolled for less than half-time. For example, the Higher Education Amendments of 1986 (Public Law 99-498) changed the definition of independent student starting with award year 1987-88. Students are now classified as independent if they Page 34 GAO/HRD90-73BR Estimating Pell GrantProgram Costs Appendix II Department of Education’s Process for Estimating PeU Grant Program Costs state, among other things, that they are either over 23 years old, a vet- eran of the Armed Forces, an orphan or ward of the court, or had legal dependents other than a spouse. Because the model could not appropri- ately simulate the effect of these changes, the Department adjusted the model estimate to assume the number of recipients affected by the new definition. The Department further adjusts its budget estimate to reflect differ- Budgetary ences in available funding and program costs from previous and current Adjustments year’s estimates. These changes are considered budgetary adjustments. The previous years’ changes are for the 2 fiscal years preceding the budget year. The 2 prior fiscal years are used because, at the time the initial budget estimate is prepared, data from the current fiscal year have not been finalized and the subsequent fiscal year is about to begin. The Department begins to prepare its budget estimate about 2 years in advance of the start of the award year. This step also includes adjust- ments made to reflect estimated savings from the Department’s legisla- tive proposals. The Department’s initial fiscal year 1990 budget estimate was adjusted for expected shortages of $96 million for fiscal years 1988 and 1989. Page 36 CiAO/IUWQO-73BB Estimating Pell Grant Program Costa Appendix III Table Supporting Figures in Report Text Table 111.1:Comparison of Pell Grant Cost Model Estimates, Actual Program Costs, Dollars in bilhons and Final Budget Requests: Award Years ~ .-- ~___~ ~~ ~~ ~- Actual program Final budget 1980-81 to 1988-89 Award year Model estimate cost request 1980-81 $2.464 $2.415 $2.444 1981-82 ..-- - 2.422 2.299 2.820 1982-83 ~ ~..-~ 2.433 2.430 2.188 1983-84 2.447 2.806 1.400 1984-85 3019 3.048 2.664 1985-86 3.681 3.586 2.800 1986-87 3.576 3.455 2.691 1987-88 3.862 3.754 3.250 1988-89 4.519 4.468 2.710 Page36 GAO/HRD-00.73BR Estimathg Pell GrantProgramCoata Appendix IV Department of Education’s Fiscal Year 1990 Pell Grant Program Budget Request The Department began to prepare its fiscal year 1990 budget request in the summer of 1988, about 2 years in advance of the 1990-91 award year. The fell grant program budget is part of the federal budget sent to the Congress in January of each year. However, it is prepared for an I award year that begins 9 months after the start of the fiscal year. The Department normally submits one annual Pell grant program budget request to the Congress. However, for fiscal year 1990 it prepared four budget estimates-an initial budget request and three revisions-the last of which was the June 1989 budget request. Two revisions resulted s in lower budget estimates while the third was $522 million higher than its original January estimate. Table IV.1 compares the major cost com- ponents of the four estimates. I Table IV.1 : Comparison of Education’s Fiscal Year 1990 Pell Grant Budget Dollars in mllhons Estimates Jan. 1969 Feb. 1989 Apr. 1989 June 1989 Model estimate $4,274.5 $4,222.2 $4,205.2 $4,835.2 Program adjustment9 369 7 366.8 366.8 270.3 Budget adjustments 95 8 3.0 -55.0 156.5 Total budget estimate $4,740.0 $4,592.0 $4,517.0 $5,262.0 %cludes admlnlstrative allowance of $5 per recipient paid to schools The Department’s June 1989 budget request was $522 million more than its original request in .January 1989. Discussions follow on (1) the rea- sons why four estimates were made, (2) how each estimate differs from the previous one, and (3) an overall comparison of the January and June 1989 budget requests. The Department’s *January 1989 estimate of $4.74 billion, which pro- January 1989 jetted 3,142,OOOPell grant recipients, was based on the June 1988 eco- Estimate nomic indicators. The estimate did not incorporate the available November 1988 economic indicators because Department officials said they did not have time to update the computer model to reflect them. (The Office of Management. and Budget issues economic indicators twice a year-in June and November-for agencies to use in developing their initial and mid-year budget estimates.) Page 37 GAO/HRD-90-73BR Estimating Pell Grant Program Costs Appendix IV Department of Education’s Fiscal Year 1990 Pell Grant Program Budget Rquest February 1989 administration’s budget. This estimate projected 37,000 fewer recipients Estimate and a decrease of $148 million from the January estimate. Contributing to the decline were a (1) $93 million decrease in the budgetary adjust- ment for an anticipated fund shortage (from $96 million to $3 million); (2) $52 million decrease in the computer cost model estimate, which was based on the November 1988 economic indicators; and (3) $3 million decline in manual program adjustments. While the February estimate was being processed at the Office of Man- agement and Budget, the American Council on Education discovered an error in the computer model. Advised of the error, the Department cor- rected its model and prepared another estimate in April. The Office of Management and Budget decided to wait until its mid-session review in June to process a budget amendment incorporating the changes. 1 In April 1989 the Department revised its budget estimate because of the April 1989 Estimate model error. The new estimate decreased $75 million from the February estimate primarily because of (I) a reduction of $17 million in the com- puter cost model estimate which was attributable to a programming error, and (2) a $58 million budgetary adjustment from an anticipated shortage of $3 million to an expected surplus of $55 million. The Department’s June 1989 budget estimate increased $745 million June 1989 Estimate from its April estimate and included a projected increase of 380,000 grant recipients. The $745 million increase consisted of (1) a $630 mil- lion increase in the cost model estimate, (2) program adjustments which reduced the estimate by $97 million, and (3) a net budgetary adjustment for an anticipated shortage of $212 million (which included the Depart- ment’s proposed legislative program changes that would save $174 million). The cost model estimate increase was due in large part to (1) a growth in applications, (2) an increase in the number of independent students, and (3) a requirement that applicants for Stafford loans and Supplemental Loans for Students apply first for Pell grants. The program adjustment reduction was largely due to a $104 million decline in the estimated cost attributed to financial aid administrators’ discretion. The $174 million in savings was based on the Department’s legislative proposals requiring (1) students admitted to postsecondary institutions Page 38 GAO/HRD90-73BR Estimating Pell Grant Program Costs Appendix IV Department of Education’s Fiscal Year 1990 Pell Grant Program Budget Request without a high school diploma to pass a test, measuring their ability to I benefit from the education, prior to enrollment, and (2) grant recipients who generally fail to complete their periods of enrollment to forfeit a pro rata portion of their grants. This latter proposal, projected to save $66 million, was included in the Department’s fiscal year 1990 appropri- I ation act (Public Law 101-166). The Department’s June budget request projected an increase of about Comparison of 842,000 more applicants and almost 500,000 more recipients than its January and June January request. Department officials said the reasons for the increases Budget Requests were based on their analysis of operational data that showed consistent increases in both over comparable periods of time. Table IV.2 shows the number of expected applicants and recipients and the economic indica- tors the Department used in developing the two estimates. Table IV.2: Comparison of Award Year 1990 Pell Grant Applicants and January June Recipients in the Department’s January request ~...~.-- request -.Change and June 1989 Budget Requests Applicants 6,310,350 7,151,944 +El41,594 Recipients 3,142,ooo 3,640,895 +498,895 Percentage receivrng awards 50 51 +1 Date of economic Indicators June 1988 June 1989 1 Year In comparing the cost elements of the January and June budget requests we found that the increase in the June request consisted of (1) a higher computer cost model estimate ($560.7 million), (2) an increase in the expected cost for grants to students attending school less than half-time ($5 million), (3) a higher administrative allowance to schools due to an expected increase of 500,000 recipients ($2.5 million), and (4) a budget- ary adjustment of $60.7 million. These increases were partially offset by a $106.9 million decrease in the estimate of financial aid administrator’s discretion and projected savings of $174 million. Table IV.3 analyzes the changes in the major cost components of the two estimates. Page 39 GAO/HRD-90.73BR Estimating Pell Grant Program Costs Appendix N Department of Education’s Fiscal Year 1990 Fell Grant Program Budget Request Table IV.3 Major Cost Components for the January and June 1989 Pell Grant Dollars in millions -.-~ Budget Requests January June Increase/ request request decrease Model estimate __~.-~. ~. $4,274.5 -_ __ $4,835 2 $560.7 Program adjustments -. - independent .-- students ~~__ .~_____I 152.0 1520 0.0 Less-than-half-time .- students 38.0 -- 43.0 5.0 -___- Financial aid administrators’ discretion ...- 164.0 57.1 -106.9 ___~ AdmInistrative allowance 15.7. 18.2 2.5 Budgetary adjustments 95.8 156.5” 60.7 Total $4,740.0 $5,262.0 $522.0 aThis figure Includes the estimated savings of $174 million for the Department’s proposed legislative changes (The Department agreed that we could consider these savings as a budgetary adjustment,) Without these savings the difference between the January and June budget requests would have been $696 mllllon, rather than $522 mhlion. Page 40 GAO/HRD90-73BR Estimating PeB Grant Program Costs Appendix V Major Contributors to This Briefing Report Joseph J. Eglin, Assistant Director (202) 245-9623 Human Resources Christopher C. Crissman, Assignment Manager Division, Theodore F. Zeunges, Evaluator-in-Charge Washington, DC. Page 41 GAO/HRD90-73BR Estimating Pell Grant Program Costs i i Page 42 GAO/HRD90-73BR Estimating Pell Grant Program Costs i i Page 43 GAO/HRD90-73BR Estimating Pell Grant Program Costs Related GAO Products Pell Grants: Who Receives Them and What Would Larger Grants Cost June 14, 1988) (GAO/HRD-~~-~~~BR, GAO'SViews on the Default Task Force’s Recommendations for Reducing Default Costs in the Guaranteed Student Loan Program (GAO/T-HRD8%7, Feb. 2, 1988) Pell Grant Validation Imposes Some Costs and Does Not Greatly Reduce Award Errors: New Strategies are Needed (GAO/PEMD-85-10,Sept. 27, 1985) Many Proprietary Schools Do Not Comply with Department of Educa- tion’s Pell Grant Program Requirements (GAO/HRD-84-17, Aug+ 20, 1984) (104637) Page 44 GAO/BRDQO-73BR Estimating Pell Grant Promm Costs . ,. : ; ii : ., :. ..I ,. -. , _’ : .’ .I . “.
Pell Grants: How the Department of Education Estimates Program Costs
Published by the Government Accountability Office on 1990-02-21.
Below is a raw (and likely hideous) rendition of the original report. (PDF)