Private Grant
Discounting tuition towards disaster. Study the effects of college tuition discounting.
Findings indicated that: 1. At least 1/4 of the colleges and universities used discounting strategies that resulted in large losses of tuition revenue 2. Institutions with the greatest increases in discount rates raised their spending on institutional grants by $3,375 per undergraduate, but their tuition and fee revenue grew by just $3,069; 3. Discounting strategies do not appear to have significantly improved the academic profiles of admitted undergraduates when measured by changes in median admissions test scores of entering first-year students 4. Tuition discounting appear to have helped institutions increase their number of low-income undergraduates 5. Increased use of tuition discounting appear to have made it possible for more students from all income levels to enter higher education.

Colleges and universities have several distinct goals for using tuition discounts: * to increase enrollments of low-income and other under-represented students; * to raise enrollments of students with high academic achievements or other talents; * To increase revenue from tuition and fees.
However, the results show that at least one quarter of the four-year private colleges and universities used discounting strategies that resulted in large losses of tuition revenue.
The institutions with the greatest increases in discount rates raised their spending on institutional grants by $3,375 per full-time equivalent (FTE) undergraduate, but their tuition and fee revenue grew by just $3,069. Thus, these institutions lost at least $306 per FTE in net tuition revenue as a result of their increases in spending on tuition discounting. Some highly selective and selective institutions lost more than $800 per FTE in tuition revenue. Due to the large losses in revenue, these institutions had smaller increases in the funds they devoted to academic instruction and other educational services to students, and had declines in spending on maintenance of campus buildings and other facilities. The institutions with the largest increases in spending on tuition discounting also saw their total undergraduate enrollments decline by 5 percent and had lower six-year graduation rates. These results occurred because discounting strategies are often focused more on increasing enrollments of first-year students than on retaining students toward graduation.
Discounting strategies also do not appear to have significantly affected the academic "profiles" of admitted undergraduates, when measured by changes in median admissions test scores of entering first-year students. Data from the College Board show that the median composite Scholastic Aptitude Test (SAT) scores of first-year undergraduates who entered college from 1990-91 to 1997-98 grew by less than 3 percent at institutions with the largest increases in tuition discount rates. In the same period, the scores of freshmen who enrolled at institutions with the smallest increases in discounting grew by 10 percent. Among colleges and universities with less-selective admissions criteria, median SAT scores fell by 2 percent at institutions with the largest increase in tuition discount races.
However, tuition discounting does appear to have helped institutions to increase their numbers of low-income undergraduates. The number of students who received Federal Pell Grants (grant aid targeted toward students with the greatest financial need) grew by 20 percent at institutions with the largest increases in discount rates, and by 16 percent at colleges and universities with the smallest increases in discounting. These results suggest that tuition discounting was more successful at helping institutions achieve their goals of providing greater access to higher education for low-income students than at attracting more academically-talented freshmen. Unfortunately, the rapid use of tuition discounting led to large losses in net tuition revenue and have resulted in decreased spending on instruction and other services to students.

WHY INCREASED: Institutional aid increased in part because the amounts provided by the federal government for Pell Grants and other grant-based assistance to low- and moderate-income undergraduates fell by nearly 6 percent, while tuition and fee charges at four-year private colleges and universities increased by 16 percent (College Board 1999a and 1999b). In more recent years, increases in federal Pell Grant appropriations have given more grant funds to low-income students, but these increases have not kept pace with rising college tuition and other charges. Consequently, many private colleges have had to increase grant assistance from their own resources in order to provide educational opportunities to low-income undergraduates. At the same time, many private institutions felt compelled to increase grant aid not only to low-income undergraduates, but also to middle- and upper-income students. Congressional panelists, in their recent study of rising college expenses (National Commission on the Cost of Higher Education 1998), expressed concern that college was becoming unaffordable for many students from higher-income families. Due to changes in the formulas used to determine eligibility for financial aid, more students from middle-and upper-income families became eligible to receive assistance. Since relatively fewer federal grant dollars were available for these students, institutions became the primary source of grant aid for them.
Many higher education analysts also became concerned about the amounts of federal student-loan debt middle- and upper-income students incurred to receive bachelor's degrees. Borrowing by students from higher-income families more than tripled from 1992-93 to 1995-96. As a result, cumulative federal student-loan debt for students from families with income of $60,000 to $79,999 grew by 43 percent, and debt for students from families with income of $80,000 or more rose by 50 percent (Redd 1999). Higher education institutions and particularly private colleges and universities increasingly felt pressured to provide more grant aid in order to lower students' loan-debt burdens.
In addition, competition among colleges for students with high academic ability is particularly fierce. Methodologies used by US. News and World Report and other organizations to rank colleges and universities usually place a high emphasis on the proportion of entering first-year students with high college-admissions test scores and other demonstrated abilities (Morse and Flanigan 2000). Accordingly, many four-year private institutions have begun to use more of their institutional aid dollars to entice the "best and brightest" students to enroll on their campuses.
Also, other universities in order to improve marketing and enrollment management strategies, decided to give tuition discounts. An increasing number of colleges and universities use their institutional aid funds to encourage students with high academic achievement or other talents to attend their institutions. Aid awarded for these efforts is also referred to as "non-need-based" grants since, in many cases, students' financial need is not a criterion for distributing these dollars. Institutions that provide non-need-based aid hope to enhance their enrollment management goals. Such goals vary; some colleges and universities simply want to increase total enrollments, while others want to raise their enrollments of students with particular abilities or talents (for example, students with admissions test scores above a certain level). Both goals have important effects on institutions' decisions to distribute institutional aid. In order to meet educational equity goals, colleges and universities must give more grants to low- and moderate-income students whose families could not afford to pay the full the tuition price. At the same time, institutions may need to give more aid to meritorious students, since these students may raise the "profile" of the institutions in the eyes of the general public and college ranking organizations.
Simultaneously, campus administrators also hope that providing large institutional grants to more students will, in the long run, bring more revenue to their colleges and universities. Many administrators believe that providing the discounts to students who pay part of the tuition and fees is better than having empty classroom and dormitory space, which generates no new revenue (McPherson and Schapiro 1998). Institutional grant dollars, if spent strategically and wisely, can help increase revenue from tuition and fees, and might raise total enrollments to levels above what they would have been had no aid been provided (McPherson and Schapiro 1998; Baum 2000). They hope that increases in grant aid, in combination with other factors (such as campus location and climate, academic reputation, etc.) will encourage more students to choose their institutions. As tuition and fee prices increase, the administrators assume that students and their families will choose the college or university that offers the lowest out-of-pocket cost or largest amount of institutional grants (Winston and Zimmerman 2000).
Ultimately, the institutions hope that the amounts they provide for tuition discounts are large enough that more students will choose to enroll, but not so large that they result in a loss of tuition and fee revenue. To meet these enrollment and revenue objectives, colleges and universities seek to set an appropriate tuition discount rate. This rate is based on the percentage of tuition and fee revenue used to cover institutional grant aid. This percentage is calculated by dividing gross tuition and fee revenue (revenue before the value of tuition discounts is subtracted) by the amounts of funded and unfunded tuition discounts. As discount rates increase, institutions run the risk of having fewer dollars available for instruction, libraries, faculty and administrators' salaries, and ocher academically related expenses. From 1989-90 to 1995-96, the tuition discount rate at all private colleges and universities rose from 17 percent to 22 percent (U.S. Department of Education 1990aand 1998a).
Awarding tuition discounts brings many potential conflicts for colleges and universities. Institutions that increase their amounts of non-need-based aid risk losing their ability to enroll more lower-income students because students who meet the non-need criteria are more likely to come from middle- and upper-income families (Baum2000; Heller and Nelson Laird 1999). At the same time, targeting more aid on low-income students might lead to a loss of tuition dollars because these undergraduates usually require larger discounts than do students from middle- and upper-income families. And providing large discounts to any number of students, regardless of the purposes, might lead to a loss of revenue for a number of institutions, particularly small colleges that do not have large resources to bear the costs of increasing grant aid.
The increasing use of tuition discounting has raised several questions that will be addressed in this study: * Has the increasing use of tuition discounting helped to raise tuition revenue at private colleges and universities? * What effect has discounting had on enrollments of low-income students, particularly at private colleges and universities with selective and highly selective admissions criteria? * Has the use of discounting increased the academic "profile" of admitted and enrolled students, as measured by admission test scores?
Study Methodology
To study the effects of tuition discounting, the USA Group Foundation obtained a database of 275 four-year private colleges and universities that responded to the Institutional Student Aid Survey sponsored by the National Association of College and University Business Officers (NACUBO 2000). These institutions responded annually to the NACUBO survey from 1990-91 to 1998-99. The database serves as a sample of the 1,463 accredited four-year private colleges and universities in the United States during the 1990s (Reindland Redd 1999).
The NACUBO data provide the most up-to-date information available on institutional grants provided by four-year private institutions. The NACUBO data include grants provided to undergraduates exclusively. These grants were funded by both tuition and fee revenue and earnings from endowments. The grant figures also include unfunded tuition waivers. The NACUBO database also provides information on the colleges and universities' annual tuition and fee charges, numbers of first-time, full-time freshmen, and numbers of freshmen who received institutional grants.
The NACUBO data were matched with information from the U.S. Department of Education's Integrated Postsecondary Education Data System (IPEDS) fall enrollment surveys (U.S. Department of Education 19906 and 1997a) in order to compare the changes in tuition discounting with the changes in total undergraduate enrollments. The most recent IPEDS enrollment data are for fall 1997 (the beginning of the 1997-98 academic year). It is likely that it would take up to one year for changes in institutions' tuition discounting policies to influence prospective students' enrollment decisions. For this reason, the NACUBO data for 1990-91 to 1996-97 were used to show what effects discounting may have had on enrollments in 1997-98. Nine institutions with missing IPEDS enrollment data were excluded from the study.
Data for low-income students are based on the number of undergraduates who received Federal Pell Grants in 1990-91 and 1997-98 (U.S. Department of Education 1991a and19986). The Pell Grant program provides funds to undergraduates from low-income families who demonstrate great need for financial assistance to attend postsecondary education. The maximum award for all students was $3,125 in academic year 1999-2000(U.S. Department of Education 2000). In 1995-96, nearly 88 percent of the Pell Grant recipients who attended four-year private colleges and universities came from families with less than $30,000 in adjusted gross income (U.S. Department of Education 1999).Changes in the number of Pell Grant awardees at private institutions thus serve as a useful indicator of trends in the enrollments of students from low-income families.
Changes in the use of tuition discounting may differ because of the criteria used by colleges and universities to admit undergraduate students. Large universities with strict admissions standards would probably award aid in a much different manner than would small liberal arts colleges with less restrictive admissions methodologies. For this reason, the colleges and universities in the database were divided into three groups of undergraduate admissions selectivity: 1. Highly selective colleges and universities that offer admission to 30 percent or less or their applicants. Just 19 institutions (or about 7 percent) meet this definition. 2. Selective institutions that offer admission to up to 60 percent of their applicants. Forty-eight colleges and universities (or 18 percent) meet this standard. 3. Less-than-selective colleges and universities that provide admissions offers to 60 percent or more of their prospective students. There were 193 institutions (73percent) in this group.
Recent Trends in Tuition Discounting,Enrollments, and Pell Grant Recipients ue to the factors previously cited, tuition discount rates increased rapidly at all admissionsselectivity levels, particularly at selective and less-than-selective institutions. Between 1990-91 and 1996-97, the average discount rate at less selective colleges and universities jumpedfrom 21 percent to 31 percent. At selective institutions, the average rate climbed from 22percent to 31 percent (see Figure 1).
In just six years, the average discount rate for all institutions rose by 9 percentagepoints, from 21 percent to 30 percent. But this increase masks some very wide variancesin discount rate changes among the institutions. Colleges and universities in the topquartile of the distribution of discount rates changesthose with "above-average" increases in discountingsawtheir tuition discount rates jump by 13.1percentage points or more. Those in thebottom quartileinstitutions with"below-average" increases in discount rates saw their rates change by just 2.5 points or less. Colleges and universities with22%21%21%
"average" increases in discount rates had growth rates of between 2:5 and 13.1percentage points. Among selective andhighly selective colleges and universities,"above-average" institutions increased theirdiscount rates by 10.9 percentage points ormore; "below-average" institutions changedtheir rates by 2.1 points or less; and"average" increases occurred between 2.1and 10.9 points. For less selective collegesand universities, "above-average" discountrates grew by 13.6 percentage points ormore; "average" institutions had rate changes of between 2.6 and 13.6 points; and "below-average" institutions had rates that changed by 2.6 points or less.
Discount rates grew rapidly because the amounts provided for institutional grantsgrew at substantially higher rates than tuition and fee revenue. Figure 2 (on the next page)displays the changes in tuition and fee revenue, tuition prices, and institutional grantspending during the study period. Gross tuition and fee revenue at selective colleges anduniversities increased by 22 percent (adjusted for inflation) per full-time equivalent (FTE)undergraduate student, but institutional grant spending per FTE jumped by 74 percent.At less selective schools, grant aid nearly doubled, but revenue grew by just 32 percent.These findings suggest that net tuition and fee revenue (gross revenue minus institutionalaid expenditures) declined by large amounts at many institutions.
Spending on tuition discounts also increased at a much higher rate than did tuition and fee prices set by colleges and universities. Tuition prices at private colleges and
Figure 1. Tuition Discount Rates at Four-Year PrivateColleges and Universities, by Undergraduate AdmissionsSelectivity Level, 1990-91 and 1996 -91
31%
31%30%
24%
15
Highly Selective
N=19
Selective
N=48
Less than SelectiveN=I93
' Includes institutions whose selectivity levels were missing.
All Institutions'
N=268III 1990-91
1111 1996-97
Source: National Association of College and University Business Officers 2000; Peterson's 1998.
11

universities rose by 21 percent in inflation-adjusted value during the study period, but therate of increase varied by selectivity level. When adjusted for inflation and weighted byfull-time equivalent enrollment, tuition and fee charges grew by 17 percent at highlyselective institutions, 20 percent at selective schools, and 23 percent at less-than-selectivecolleges (see Table 1).3
The large difference in growth rates between institutional grants and tuition priceseffectively means that many undergraduates were paying much less than the increases inlisted tuition "sticker so 45
Figure 2. Percentage Change in Institutional Aid Compared WithChange in Tuition and Fee Revenue and Tuition and Fee Charges,1990-91 to 1996-97, at Four-Year Private Colleges and Universities
96%
Highly SelectiveSelective
Less than Selective
All InstitutionsEl Institutional Aid/FTE
Tuition and Fee Revenue /FTE'7 Tuition and Fee Charges/FTE* nstitutional grants, tuition and ice revenue, and tuition and fee charges arc divided by full-time equivalent (FTE) undergraduateenrollments in Fall 1990 and Fall 1996."Includes institutions whose selectivity levels were missing.
Source: National-Association of College and University Business Officers 2000; Peterson's 1998.
12
prices" to attend four- year private collegesand universities. Thedifference between thelisted tuition chargesandtheamountsstudents and familiesactually paid is betterunderstoodbycomparing the growthinaverage tuitioncharges with averageinstitutional grants onaper-dollar basis.From1990-91to
1996-97, the averagetuition and fee price atprivate colleges anduniversities increasedby$2,684,from$12,526 to $15,210,whileaverageinstitutional grants grew by $1,835, from $2,313 to $4,148. Using a per-dollar basis tocompare these two figures reveals that for every dollar the institutions increased theirprices, they provided, on average, 68 cents in tuition discounts for each FTE undergraduatestudent. The typical undergraduate and her family actually paid only 32 cents of eachdollar of tuition increase. At selective institutions, tuition discounts essentially covered 70cents of each dollar of tuition price increase; at less selective institutions, discounts paidfor 74 cents of each dollar of tuition increase. Understanding the difference between thelisted tuition prices and the amounts actually paid by undergraduates is important becausemany policy makers and college students have become concerned about increases in tuitionand fee prices at all postsecondary education institutions (National Commission on theCost of Higher Education 1998). Ironically, these concerns grew despite the fact thatmany private colleges and universities raised their own grant spending to cover the majorityof these price increases for many students.
Tuition and fce prices grew by 15 percent at all four-year private colleges and universities. Sec College
Board 1999h.
/6BEST COPY AVAiLABLE

Table 1. Tuition and Fee Charges and Institutional Grants at Four-Year Private Colleges andUniversities, by Undergraduate Admissions Selectivity Level, 1990-91 and 1996-97
AverageAverageTuition andTuition andFee ChargesFee Chargesin 1990-91*in 1996797*
InstitutionalGrantsPer PTEPercentageUndergraduateChangein,1990-91
InstitutionalGrantsPer FTEUndergraduatePercentagein 1996-97Change
Percentage
of the Increasein Tuition and FeesCovered by theIncrease in Grants
Highly Selective$17,414$20,44417.4%$3,442$4,72437.2%':42.3%Selective15,70218,86220.1%3,0205,24773.7%-°.g0.5%Less than Selective10,6763,13323.0%.1;902.3,72195.6%74.0%All Institutions"$12,526$15,21021.4%$2,313$4,14879.3%,.;618 4%Tuition and fee charges for 1990-91 are in 1996-97 dollars. The tuition charges for 1990-91 and 1996-97 are weighted by full-time equivalent (FM)undergraduate enrollment.."Includes institutions whose selectivity levels were missing.Sources National Assodadon of College and University Business Officers 2000; Research Associates of Washington 1999...-
Much of the added grant spending by private colleges and universities appears tohave been directed toward attracting new first-year students. As Table 2 illustrates, theproportion of first-time, full-time freshmen who received grants at four-year privateinstitutions grew from about 64 percent in 1990 to 76 percent in 1996. The largestincrease occurred at less selective institutions, where in just six years the proportion of
Table 2. Percentage of First-Year Undergraduates Who ReceivedInstitutional Grants and Average Grant Awards, 1990-91 and 1996-97
Percent ofPercent ofFreshmen WhoFreshmen WhoAverageAveragePercentageReceivedReceivedGiants toGrants toChange inInstitutionalInstitutionalFt isshmenFreshmenAverageGrants in 1990-91Grants in 1996-97in 1990-91"in 1996-97GrantsHighly Selective42%.45%; .410,307Selective51%64%.,8,997Less than Selective69%83%4;353All Institution?6447646$ 5,6i2'Includes institutions whose selectivity levels were missing.
"In 1996-97 dollars.
$12,11417.5%10,44416.1%5,73331.7%$. 7,02225.1%Source: National Association of College and University Business Officers 2000; Research Associates of Washington 1999.
7 7*".,;*.^.
BEST OPIrAVA11
13
LE

aided freshmen jumped from 69 percent to 83 percent. Conversely, highly selective collegesand universities saw just a 3-percentage point gain, from 42 percent to 45 percent.
Average institutional grants to first-year students also rose substantially. Less selectiveinstitutions increased their average grant award to first-year students by 32 percent ininflation-adjusted value, from $4,354 to $5,722, while grants to those at highly selectivecolleges rose by 17
Figure 3. Percentage Change in Undergraduate Enrollments and PellGrant Recipients* at Four-Year Private Colleges and Universities, byUndergraduate Admissions Selectivity Level, 1990-91 and 1997-98
30%Total Enrollment
First-Time Freshmen22%
3Pell Grant Recipients
Highly Selective
SelectiveLess than SelectiveAll Institutions"
'MI Grant, total enrollment, and first-time freshmen enrollment data are for 1990-91 to 1997-98. Study assumes a one -year lagperiod between changes in discount rates and changes in enrollments and Pell Grant recipients.**Includes institutions whose selectivity levels wcrc missing.
Source: National Association of College and University Business Officers 2000: Peterson's 1998; U.S. Department of Education1990h. 1991a, I 997a. and 1998b.
14
percent, from $10,307to $12,114. Collegesand universities with thegreatest increases indiscount rates also hadthe largest growth ininstitutional grants tofreshmen.Averageinstitutional grants tofirst-year undergrad-uates at colleges anduniversities with above-averagegrowthindiscount rates jumpedby 52 percent. Averagegrants to freshmen atinstitutions with below-average and averagerates of change indiscount rates grew byjust 21 percent and 28percent, respectively.Despite these largeincreases, many collegesand universities appar-ently struggled to meet their enrollment management goals. Wolff and Bryant (1999)report that the average yield rates (percentage of admitted students who actually enrolled)at private colleges and universities actually fell from 44 percent in 1995 to 38 percent in1997.
The additional grant aid to freshmen helped to increase the number of first-yearundergraduates at all selectivity levels, but total undergraduate enrollments at mostinstitutions were essentially flat, as Figure 3 demonstrates. Between 1990-91 and 1997-98, the total number of first-time, full-time freshmen at four-year private colleges anduniversities grew by 10 percent, but total undergraduate enrollment increased by just 2percent. Even though freshmen enrollment at selective colleges and universities increasedby 9 percent, total enrollment dropped by about 1 percent. Highly selective and less-than-selective schools had total enrollment increases of between 1 percent and 2 percent, whilefreshmen enrollment was up 5 percent and 11 percent, respectively.
Figure 3 also shows that colleges and universities at all selectivity levels had some success in increasing their numbers of low-income undergraduates. At selective and highly
/9
EST COPYMAKE

selective institutions, the number of Pell Grant recipients grew by 25 percent and 30percent, respectively, while the number of those who received Pell Grants was up 20percent at less selective colleges and universities. In spite of these large increases at selectiveand highly selective institutions, Pell Grant recipients are still over-represented at lessselective colleges and universities. In 1997-98, nearly three quarters of those who got PellGrants to attend private colleges and universities were enrolled at less-than-selectiveinstitutions. The percentage of undergraduates who received Pell Grants at highly selectiveinstitutions grew only slightly, rising from about 9 percent to 12 percent. At selectivecolleges and universities, the proportion of students who received Pell Grants grew from14 percent to 18 percent, and the share of undergraduates at less selective institutionswho got Pell Grants increased from 21 percent to 24 percent.
In addition, the income levels of Pell Grant awardees changed during the period. In1990-91, 60 percent of Pell Grants recipients came from families with incomes of $15,000or less. By 1997-98, the proportionof recipients from this income groupdeclined to just 43 percent (seeFigure 4). In the same period, theshare of recipients from families withincomes of more than $30,000 grewfrom 9 percent to 23 percent. As aresult, the median family income ofPell Grant recipients jumped byalmost 59 percentfrom $1,386to $18,088 (U.S. Department ofEducation 19916 and 1998b).Several factors might account for thisincrease. The rising costs of collegevery likely made it possible for morestudents from relatively higher-
Figure 4. Family Income Levels of Pell Grant Recipients at Four-Year Private Colleges and Universities,1990-91 and 1997-98
6060%
43%
34%
3030%
15%
7% income families to become eligible815,000 & Under$15,001 to 530.000for Pell Grants. Further, Gladieuxand Swail (1999) show thatSource: U.S. Department of Education I 99Ib and I998c. throughout the 1990s students fromfamilies with the lowest incomes were the least likely to attend postsecondary education.In 1996, just 54 percent of high school graduates from families in the lowest incomequartile ($24,589 or less) were attending postsecondary education, compared with 85percent of those from the highest income quartile ($71,802 and over). Still, throughoutthe study period, more than three quarters of the Pell Grant recipients came from familieswith incomes of $30,000 or less.
530.001 to 540.000
19
8%
$40.000 & Over1111 1990.91
II 1997-98
BEST COPYAVAILABLE
15

For at least one quarterof the four-year privateinstitutions, discounting has resulted in asubstantial loss in nettuition revenue.
16
Tuition Discounting Compared WithNet Tuition Revenue
-.egardless of their goals for using tuition discounting, institutions have one standard thatcan be used to determine how their increases in grant spending affected their tuition andfee revenue: Was the marginal tuition and fee revenue generated by discounting greaterthan the marginal costs of providing this aid (Breneman 1994)? Or, put much more simply,were institutions able to increase their tuition revenue as a result of discounting? Were theincreases in revenue greater than the costs of providing the discounts?
Marginal cost, marginal revenue, and net revenue figures were calculated to answerthese questions. Marginal cost is simply the difference between the amounts provided forinstitutional grants in 1990-91 and 1996-97, while marginal revenue is the difference ingross tuition and fee funds collected by the institutions during the same period. Net revenueequals the difference between marginal revenue and marginal cost. All three figures weredivided by full-time equivalent undergraduate enrollment to adjust for changes in numbersof students. The figures also were adjusted for inflation.
At the very least, colleges and universities should ensure that net revenue is equal tozero, which would mean that their costs of discounting were no greater than the additionalrevenue they generated from having more students enroll on their campuses. Optimally,net tuition revenue should be greater than zero. If net revenue were greater than zero, thenmore funds would be available for educational and other programs on campus.
This is not to imply that institutions use discounting to "make a profit." On thecontrary, these private colleges and universities are non-profit educational institutionsthat would be expected to use any additional funds to develop and expand academic andother programs that ultimately benefit students. But colleges and universities should not"lose" money because of discounting, for such losses could eventually harm academicprograms and make it more difficult to provide quality education to their students.
At this point, a distinction should be made between a "paper" and a "real" loss in nettuition revenue that might result from discounting. Institutions with large endowmentsusually can fund their discounts with endowment earnings. Thus, if one were to comparetuition and fee revenue with increases in grants funded by tuition revenue exclusively, "onpaper" these institutions might appear to have lost net revenue. In real terms, once fundsfrom endowments were added, these institutions might have gains in net tuition revenue(or their losses would be much smaller). NACUBO's database includes discounts fundedby both tuition revenue and endowment earnings, so any gains or losses in revenue are in"real" terms.
Based on the standards stated above, it appears that, for at least one quarter of thefour-year private institutions, discounting has resulted in a substantial loss in net tuitionrevenue. As Table 3 demonstrates, at institutions with above-average increases in discountrates, marginal spending on institutional grants rose by $3,375 per FTE undergraduate.Conversely, marginal tuition and fee revenue increased by just $3,069. The institutionswith above average growth in tuition discount rates lost $306 per FTE as a result of theirincreases in spending on institutional grants.
Another way to look at these figures is to divide the marginal increase in grant spendingby the marginal increase in revenue, which would equal the percentage of marginal revenue
20
BM COPY MAILABLE

used by institutions to cover their marginal increases in grants. The above-averageinstitutions used 110 percent of their marginal increase in revenue to fund their increasesin tuition discounting. In other words, for each dollar these institutions gained in marginaltuition revenue, they provided $1.10 in tuition discounts. Essentially, some undergraduates
Table 3. Marginal Increase in Institutional Grants, Marginal Increase in Tuition andFee Revenue, and Net Tuition and Fee Revenue for Four-Year Private Colleges andUniversities, by Level of Change in Tuition Discount Rates, 1990-91 and 1996-97
MarginalMarginalNet RevenuePercent ofPercentage PointIncrease inIncrease in(Marginal Rev.MarginalChange in TuitionNumber ofInstitutionalTuition and FeeMinus MarginalRev. Used forDiscount RatesInstitutionsGrants per FTERevenue per FTECost) Per FTEAdded GrantsBelow Avg (2.5 Points or Less)66$ 637$3,481$2,84418.3%Avg (2.5 to 13.1 Points)1341,6402,9871,34754.9%Above Avg (13.1 Points or More)663,3753,069-306110.0%Total266$1,835$3,105$1,27059.1%Table 4. Marginal Increase in Institutional Grants, Tuition and Fee ReVenue, and NetTuition and Fee Revenue for Selective* Four-Year Private Colleges, by Level ofChange in Tuition Discount Rates, 1990-91 and 1996-97
Marginal1 MarginalNet RevenuePatent ofPercentage PointIncrease inIncrease in(Marginal Rev.MarginalChange in TuitionNumber ofInstitutionalTuition and FeeMinus MarginalRev. Used forDiscount RatesInstitutionsGrants per FITRevenue per FITCost) Per FITAdded GrantsBelow Avg (2.1 Points or Less)17$ 800$4,981$4,18116.1%Avg (2.1 to 10.9 Points)331,5232,9541,43151.6%Above Avg (10.9 Points or More)173,5952,762-833130.2%Total67$1,922$3,340$1,41857.5%'Includes highly selective and selective four-year private colleges and universities.Source: National Association of College and University Business Officers 2000; Research Associates of Washington 1999. were actually paying "negative tuition." The amounts some students received in discountswere greater than their institutions' listed tuition and fee prices (Winston and Zimmerman2000). Institutions that lost net tuition revenue may have been able to make up for someof these losses through increased dollars from other sources (for example, from room andhoard charges). However, given the great dependency on tuition and fee funds at manyprivate institutions, it is unlikely that any gains in fundsfrom other sources were largecnough to offset the losses in tuition revenue completely. BET
17
ON NA URI

The results were very different for institutions with below-average changes in discountrates. These institutions gained $3,481 per FTE in marginal revenue, while marginalcosts rose by just $637, fora net gain of $2,844 per FTE. Each dollar of increased tuitionrevenue "cost" these institutions just 18 cents in added tuition discounts. For all institutions,net revenue increased $1,270 per FTE, or by 41 percent. Each dollar gained in tuition andfee revenue "costs" the institutions 59 cents in added tuition discounts.
Highly selective and selective colleges and universities' with above-average increasesin discount rates had even greater losses in net revenue. Marginal grant spending at theseschools jumped by $3,595 per FTE, but marginal revenue grew by only $2,762, for a lossof $833 per FTE (see Table 4 on the previous page). For each dollar these institutionsgained in marginal revenue, they provided $1.30 in additional tuition discounts. This compares
Table 5. Change in Institutional Grants, Tuition and Fee Revenue, and Net Tuition andFee Revenue at Less-Than-Selective Four-Year Private Colleges and Universities, byLevel of Change in Tuition Discount Rages, 1990-91 and 1996-97
Marg'malMarginalNet RevenuePercent ofPercentage PointIncrease inIncrease in(Marginal Rev.MarginalChange in TuitionNumber ofInstitutionalTuition and FeeMinus MarginalRev. Used forDiscount RatesInstitutionsGrants per FITRevenue per FTECost) Per FTEAdded GrantsBelow Avg (2.6 Points or Less)48$ 525$2,776$2,25118.9%Avg (2.6 to 13.6 Poinp)971,6852,9541,26957.0%Above Avg (13.6 Points or More)483,2913,289-2100.1%Total193$1,819$2,960$1,14161.5%Source: National Association of College and University Business Officers 2000; Research Associates of Washington 1999. with a gain in net revenue of $4,181 per full-time equivalent undergraduate (or 84 percent)for selective institutions with below-average growth in discount races. On average, selectiveinstitutions gained $1,418 per FTE (43 percent) in net revenue. Thus, about 57 percentof the growth in marginal revenue was used to fund the marginal increase in grants.
Added tuition revenue at less selective colleges and universities with above-averageincreases in discount rates barely kept pace with their growth in grant spending. Theseinstitutions had a $3,289 per FTE increase in marginal tuition revenue, but grant spendinggrew by $3,291 per FTE, for a lost of $2 per FTE undergraduate. Meanwhile, collegesand universities with below-average growth in discount rates had a net revenue gain of$2,251 per full-time equivalent student. On average, the less selective colleges anduniversities had a gain in net revenue of $1,141 per FTE undergraduate, a 39 percentincrease in net revenue. These figures mean that 61 percent of thegrowth in marginal revenuewas used to And the increased spending on tuition discounts. Table 5 displays these results.
Due to the small numbers of these colleges and universities, the selective and highly selective institutionswere combined for this analysis. two t!) 18
EST COPY AVAILABLE

Ironically, the colleges and universities with above-average increases in discount rateswere the most dependent on tuition and fee revenue to finance their basic educationaloperations. In 1995-96, institutions that increased discounting by above-average ratesreceived nearly two-thirds of their educational and general revenue from tuition and fees.Tuition and fee dollars accounted for just 45 percent of the education-related funds at
Table 6. Percentage Change in Amounts Spent on Instruction, PlantMaintenance, and Academic Support at Four-Year Private Collegesand Universities, 1990-91 to 1995-96 Instruction
Amount Spentin 1990-91(in $1,000s)*
Amount Spentin 1995-96(in $1,000.)
Percentage
Change
Below Avg (Less than 2.5 Points)$1,293,110$1,540,25419.1%Avg (Between 2.5 and 13.1 Points)2,774,3163,089,95711.4%Above Avg (13.1 Points or More)984,4691,015,9983.2%All Institutions**$1,065,480$1,149,1407.9%Plant Maintenance and OperationsAmount SpentAmount Spentin 1990-91in 1995-96Percentage(in $1,000s)*(in $1,000.)Change
Below Avg (Leas than 2.5 Points)$ 269,832$ 288,572*6:9%Avg (Between 2.5 and 13.1 Points)614,346656,0616.8%Above Avg (13.1 Points or More)232,197230,392-0.8%All Institutions**$1,116,375$1,175,0255.3%Academic SupportAmount SpentAmount Spentin 1990-91in 1995-96(in $1,0005)*(in $1,000s)
PercentageChange
Below Avg (Less than 2.5 Points)$ 222,544$ 244,0669.7%Avg (Between 2.5 and 13.1 Points)625,820675,9978.0%Above Avg (13.1 Points or More)217,116229,0775.5%All Institutions**$1,065,480$1,149,1407.9%'Data are in 1995-96 dollars.
**Expenditure data for three institutions were missing.
Source: National Association of College and University Business Officers 2000; Research Associatesof Washington 1999; U.S. Department of Education 1990a and 1998a.
2319
I COPYAVMPILE

Despite their substantial growth in grantspending, institutionswith the largestincreases in discounting saw the biggestdeclines in totalundergraduateenrollments. 20 colleges and universities with below-average growthin discount rates, and for 58 percentat those with average changes in rates (U.S. Department of Education 1998c).
It is no coincidence, then, that institutions that raisedtheir tuition discounts by above-average rates also had the smallest increases inamounts spent to finance majoracademic-related campus operations. As Table 6on the previous page shows, the fundsthat institutions with above-average growth in discountingdevoted to instructional expensesincreased by only 3percent, compared with a 19 percent rise at below-average institutions.Expenditures for operation and maintenance ofcampus buildings and grounds fell by Ipercent at the colleges and universities with above-average growthrates in tuitiondiscounting, but increased by about 7percent at those with below-average and averagerate increases. And the amounts devoted to academicsupport services (expenditures forlibraries, museums,course and curriculum development, and other education-relatedservices) grew by just 5 percentat the institutions with above-average growth in discounting,compared with 10percent at colleges with below-average rate increases and 8 percent atthose with averagerates of change in discount rates.
It is quite likely that other factorswere at work on the campuses with above-averageincreases in discounting that wouldaccount for these slower increases in education-relatedexpenditures. But it seems clear that the large increasesin institutional aid spending harmedthese institutions' abilityto increase funding for some important parts of their educationaloperations. These results are notmeant to suggest that students' academic programs werecompromised by these changes in spendingpatterns, but a number of institutions apparentlystruggled with their abilityto provide larger increases in educational expenditures at thesame time they expended greater amounts for tuition discounts.
Tuition Discounting ComparedwithTotalEnrollment and Pell Grant Recipients espite then. substantial growth in grant spending, institutions with the largest increases indiscountil saw the biggest declines in totalundergraduate enrollments. The number ofundergraduates at colleges and universities with above-averageincreases in discount ratesfell by 5 percent from 1990-91to 1997-98 (see Figure 5 on the next page). In the sameperiod, undergraduate enrollmentgrew by 7 percent at the institutions with below-averagechanges in discounting, and by 3percent at institutions with average growth rates.
The decline in total enrollmentsat above-average institutions happened even thoughthese institutions increased their enrollments of first-yearundergraduates by 8 percent.However, the number of freshmenat institutions with average and below-average growthin discount rates grew by 11percent. This finding shows that the large increases ininstitutional grants did not havea strong influence on first-year students' choices whenthey decided which private institutionsto attend. Students weigh many factors outside offinancial aid (choice of academic majors, quality ofacademic instruction and facilities,campus location, etc.) when they choose to enroll at a particular collegeor university(U.S. Department of Education 1998d). Increases in institutionalgrants are not a substitutefor these other factors, and changes in discountrates by themselves cannot be expected tobe the strongest influence in students' decisionsto choose a college or university to attend.
1t
4AA
4/-
BEST COPY t WAI'

Selective institutions with above-average growth in discount rates experienced evensharper declines in undergraduate enrollments. Figure 6 displays these results. Selectiveand highly selective colleges and universities with above-average increases in discountingsaw their numbers of undergraduates fall by 8 percent, compared with a drop of less than 10
Figure 5. Percentage Change in Total Undergraduate Enrollmentsand Enrollments of First-lime, Full-Time Freshmen at Four-YearPrivate Colleges and Universities, by Level of Change in TuitionDiscount Rates, 1990-91 to 1997-98
11%11%
10%
Below Average(2.5 Points or Less)
N=66
Average(2.5 to 13.1 Points)N=134
-5%
Above Average(13.1 Points or More)N=66
All InstitutionsN=266
El Total Enrollment
U Freshmen Enrollment
Figure 6. Change in Total Undergraduate Enrollment and Enrollmentof First-Time, Full-Time Freshmen at Selective Institutions*, byLevel of Change in Tuition Discount Rates, 1990-91 to 1997-98
9%8Total Enrollment6%Freshmen Enrollment
5%
-0.2%
Below Average(2.1 Points or Less)
N=17
Average(2.1 to 10.9 Points)
N=33
-8%
Above Average(10.9 Points or More)N=17
-0.3%
All InstitutionsN=67
Includes highly selective and selective tour -year private colleges and universities.
Source: National Association of College and University Business Officers 2000: U.S. Department25 ofEduRIPIgand!997:1' 21

I percent at institutions with below-average changes in discountrates. The colleges anduniversities with above-average increases in discountrates also saw their numbers of first-time freshmen grow by only 5percent, compared with increases of 9 percent at averageinstitutions and 6 percent at below-average schools.
Figure 7. Percentage Change in Total UndergraduateEnrollment andEnrollment of First-lime, Full-lime Freshmenat Less-Than-SelectiveInstitutions, by Level of Change in Tuition DiscountRates, 1990-91 to 1997-98
1010%
14%12% II Total Enrollment
11%woFreshmen Enrollment
8%
2%2%
Below AverageAverageAll Institutions(2.6 Points or Less)(2.6 to 13.6 Points)N=193N=48N=97
-4%
Above Average(13.6 Points or More)N=48
Figure 8. Percentage Change in UndergraduateEnrollments and Pell GrantRecipients at Four-Year Private Colleges and Universities, byLevel ofChange in Tuition Discount Rates, 1990-91 to 1997-98
2010
25%
22%II Total EnrollmentII Pell Recipients
Below AverageAverage(2.5 Points or Less)(2.5 to 13.1 Points)N=66N=134
-5%
Above Average(13.1 Points or More)
N=66All Institutions
N=266
Source: National Association of College and University Business Officers2000: U.S. Department of Education,19906, 1991a, and 1997a, and 19986.
22
BEST COPYAVNILAIBLE

Results at less selective institutions were similar. Colleges and universities with above-average increases in discount rates saw their total enrollments fall by 4 percent, whileenrollments jumped by 10 percent at below-average institutions and increased by 2 percentat average institutions (see Figure 7). Freshmen enrollments grew by 8 percent at above-average institutions, but increased by 14 percent and 12 percent at colleges and universitieswith below-average and average changes in discount rates, respectively.
However, as Figure 8 demonstrates, enrollments of low-income undergraduates wereup sharply at all levels of change in tuition discount rates. For the colleges and universitieswith above-average gains in discounting, the number of Pell Grant recipients grew by 20percent, which compares favorably with the 16 percent increase in low-income enrollmentsat institutions with below-average changes in discount rates. At schools with averageincreases in discounting, enrollments of Pell Grant recipients went up by 25 percent.
These results suggest that many colleges and universities, particularly those that hadthe highest increases in discount rates, were more successful at using their institutionalgrants to attract new first-year students to their campuses, but less successful in retainingthese students toward graduation. That is, one reason total enrollment declined atinstitutions with the largest gains indiscount rates is that a greaterproportion of freshmen at thesecolleges and universities left beforereceiving their baccalaureate degrees.This situation may have caused totalenrollments to fall despite the growthin numbers of first-year under-graduates.
Table 7 illustrates this point byshowing the six-year graduation ratesfor the colleges and universities thatsupplied data to the NationalCollegiate Athletic Association(NCAA 1998a and 19986).5 Thesedata show the proportion of studentswho enrolled as degree-seeking, full-time, first-time freshmen in 1991-92,the first year the NCAA collected data,and received a bachelor's degree fromtheir institutions within six years (byAugust 1997). At colleges anduniversities with the below-averageincreases in tuition discount rates, the average graduation rate was about 70 percent,versus 63 percent at above-average colleges and universities and 67 percent at institutionswith average rates of increase in discounting. These figures show that roughly 37 percent ofthe freshmen at institutions with above-average growth in discount rates left their colleges and
Table 7. 1997 Graduation Rates* for Four-Year Private Collegesand Universities, by Undergraduate Admissions Selectivity andLevel of Change in Tuition Discount Rates
Below AverageAverageAbove Average
All Institutions**
AllInstitutions
-69.6%-67.1%63.2%
66.5%Highly Selectiveiiia7Than-and SelectiveSdictiveInstitutions'Iitutions
85.3%56.9%82.7%61.9%71.4%59.6%80.2%60.4%'Based on the proportion of first-time, full-time, degree-seeking freshmen who received bachelor's degreesfrom their original higher education' nstitutions by August 1997.
"Based on a sample of 187 four-year private colleges and universities that provided graduation rate datatothe National Collegiate Athletic Association.
Source: National Collegiate Athletic Association 1998a and 1998b; National Association of College andUniversity Business Officers 2000.
' About 70 percent (187 of 266) of the colleges and universities in the NACUBO database supplied graduationrate data to the NCAA. 2
27
BEST COPY AVAILABLE
23

Administrators thusmay need to emphasize both attractingand retainingundergraduates in orderto achieve their revenueand enrollment goals. 24 universities before receiving a bachelor's degree, compared with 30percent of those at below-average schools. A higher share of the undergraduates at institutions with above-averagegains in discount rates transferred to another collegeor university, dropped below full-time enrollment, took longer to graduate, or left higher education entirely. It is also possiblethat some colleges that award large institutionalgrants to first-year students reduce aidamounts for students who return for the second and subsequent years of higher education;these lower grant amounts may causesome students to leave their institutions after thefirst year.
At selective and highly selective institutions, theaverage graduation rate for collegesand universities with below-average changes in discountrates was about 85 percent,compared with 71 percent at colleges and universities with above-average changes. On theother hand, the 60 percent graduation rate at less selective institutions with above-averageincreases in discounting was slightly higher than the 57percent average graduation rate atbelow-average colleges and universities.
While the differences in graduation ratesare not substantial, they do suggest that arelationship exists between changes in discountrates, net revenue, and graduation rates,particularly at the more selective private colleges and universities. It is possible that totalundergraduate enrollment and net tuition revenue fellat the colleges and universities withsubstantial increases in discount rates because these institutions had higher attritionrates,a problem that campus administrators cannot solve simply by raising their institutionalaid budgets. Administrators thusmay need to emphasize both attracting and retainingundergraduates in order to achieve theirrevenue and enrollment goals.
Tuition Discounting and Changes in SATI,Scores of Admitted Freshmen. s mentioned, administrators at colleges and universities hope to use tuition discounting
.strategies to admit more students who demonstrate academic merit or other talents. Given diffIrge increase in institutionalgrants to freshmen, it is possible that schools that raisedtheir discount rates by above-average amountswere able to enroll more higher-abilitystudents. Thus, these institutions may have achieved their goal of raising the academic"profiles" of their colleges and universities.
One way to compare changes in institutions' discountrates and their admissionstandards is to look attrends in the median admissions test scores of admitted full-time, full-year undertes, by selectivity level and level of change in discount rates.The most widely used a..ssions test is the Scholastic Aptitude Test (SAT), administeredby the Educational Testing Service and the College Board. Thistest is often used to judgeprospective students' academic merit and is seen as a predictor for students' chances ofsuccess at college-level academic study. Generally, the admissions test scores of admittedstudents is an indication of institutions' level of selectivity, with the studentsat the most-selective institutions having the highestaverage scores. A substantial change in these scorescould be a strong indication of alterations in institutions' enrollment selection criteria.Median composite SAT score data (verbal and mathscores combined) for admittedfreshmen at the majority of private colleges and universities is collected by the College itNe. 2
BEST COPYMNABLE

Board's Annual Survey of Colleges (College Board, 1992 and 1999c). These data were usedto compare changes in median composite SAT scores for freshmen who enrolled in thefall of 1990' with those from the fall of 1997, by admissions selectivity level and level ofchange in tuition discount rates.
Some readers might object to the use of test scores in this way. At most four-yearprivate colleges and universities, the SAT score is just one indicator that admissions officersuse to evaluate the applications of prospective students. Others may argue that suchcomparisons are unfair to institutions that enroll large numbers of Pell Grant recipients,since students from higher-income families tend to do better on the SAT by virtue oftheir access to test-preparation courses (Chronicle of Higher Education 1999). Thus,institutions with larger numbers of Pell Grant recipients may have lower median SATscores. Still others will point out that many ocher factors besides changes in tuition discountrates would account for changes in median SAT score levels of admitted students. Forthese reasons, the use of admissions scores in this study should be considered as anapproximation of changes in institutions' selection criteria.
Based on the changes in median composite SAT scores, it appears that increases intuition discount rates have not significantly affected institutions' ability to enroll studentswith higher SAT scores. In fact, both selective and less-than-selective institutions that hadbelow-average increases in discount rates had the highest average increases in medi.SAT scores.Table 8 (on the next page) demonstrates this point. At highly selective_and selectivecolleges and universities, the median SAT score of admitted freshmen grew by nearly 10percent at institutions with below-average increases in discount rates. At colleges anduniversities with average and above-average changes in discount rates, median SAT scoresgrew by less than 3 percent. The results are even more striking at less-selective privatecolleges. At average and below-average institutions, median SAT scores actually declinedslightly, despite their sharp increases in amounts of institutional grants. Colleges anduniversities with below-average growth in discount rates had a small increase in medianSAT scores.
Selective colleges and universities may have had some success in their efforts to usetuition discounts to recruit prospective undergraduates with academic merit. AverageSAT scores for freshmen at these institutions did increase, but not by amounts largeenough to make a substantial impact at most colleges. The results were just the oppositeat less selective institutions. In fact, the small decline in SAT scores at less selective collegesand universities may suggest that more of these institutions were accepting more studentsregardless of their demonstrated academic ability in order to meet enrollment goals (Wolffand Bryant, 1999). However, these results should be judged cautiously, since none of thechanges in SAT scores are statistically significant.' It is also possible that less-selectiveinstitutions purposefully gave more grants to low-income students without regard to theirtest scores in order to give these students greater access to higher education. The findings
" In 1995, the College Board "re-centered" the SAT scoring scale to reflect changes in the test-takingpopulation. The scores for fall 1990 used in this study are based on the "re-centered" scale. For moreinformation, see the College Board's Web site, <http: / /www.collegeboard.org>.Statistical significance tests arc based on Student's T calculation with a 5 percent margin for error. qg BE
25
It appears thatincreases in tuitiondiscount rates have notsignificantly affectedinstitutions ability toenroll students withhigher SAT scores.
COPYAVAILABLE

26
Table 8. Median Composite SAT Scores for Freshmen Who EnteredFour-Year Private Colleges and Universities in Fall 1990 and Fall 1997,by Level of Change in Tuition Discount Rates
Selective and Highly Selective
MedianMedianTuitionCompositeCompositeDiscountingSAT ScoreSAT ScorePercentageChange Levelin Fall 1990*In Fa111997Change
Below Average1,1871,3029.7%Average1,2451,2802.8%Above Average1,1821,2102.3%
All Institutions1,2051,2554.1%
Less-Than-Selective
MedianMedianTuition,CompositeCompositeDiscounSAT ScoreSAT ScorePercentage in FallIn Fall 1997Chang
Below Aver4e1,0621,0700.8%
Average1,1251,095-2.7%Above Average,1,0901,080-0.9%
All Institutions1,1051,085-1.8%
'SAT Scores for the Fall of 1990 were converted to the College Board's "re-centered" SAT scoring scale.
Source: College Board 1992 and 1999c; National Association of College and University BusinessOfficers 2000. therefore suggest, but cannotprove, that there is little or no relationship between colleges'increases in discount rates and their ability to recruit higher-ability first-year students.
The findings also suggest thatsome private institutions used a substantial portion oftheir need-based institutional grantsto enroll an even larger number of low-incomefreshmen. In general, collegesappear to have been more successful at using their institutionalgrants to meet their educational equity goals, but were less successful in using tuitiondiscounts to enroll more higher-ability undergraduates.
30
BEST COPYAVAILABLE

Comparison of the NACUBO Samplewith National Data o account for any possible biases in the NACUBO sample, the study design was appliedto the entire population of 1,463 four-year private colleges and universities that providedinformation to the U.S. Department of Education's IPEDS fall enrollment and financesurveys in 1989-90, 1995-96, and 1996-97 (U.S. Department of Education 1989, 1990a,1998a, and 1998e), and Pell Grant data files for 1989-90 and 1996-97 (U.S. Departmentof Education 1990c and 1997b). By federal law,' all higher education institutions arerequired to complete the IPEDS surveys in order to remain eligible to participate in theprograms authorized under Title IV of the Higher Education Act of 1965, as amended.'IPEDS thus provides a census that can be used to assess changes in enrollments and nettuition revenue at most of the accredited four-year private colleges and universities. Amongmany other items, the IPEDS finance survey collects the amounts institutions spendannually on institutional grants and scholarships and their gross tuition and fee revenue.
Several important distinctions between the IPEDS and NACUBO databases mustbe made before these results are shown. First, the most recent year of available financedata from IPEDS is 1995-96; as mentioned, the NACUBO data are more up-to-date.Second, the IPEDS data include amounts of institutional grants provided to graduateand professional students as well as undergraduates, while NACUBO's survey includesfunds to undergraduates exclusively. While there is no way to know for sure how much ofthe aid reported by IPEDS was provided to advanced degree students, it is likely that theoverwhelming majority of the funds went to undergraduates (Reindl and Redd 1999;Heller and Nelson Laird 1999). Similarly, the tuition and fee revenue and net tuitionrevenue data from IPEDS are based on tuition amounts paid by undergraduates andgraduate/professional students. Third, the institutional aid figures reported in IPEDSmay not include funds from endowment income, so these amounts may under-reportthe total amounts of aid provided to students. As such, these data also may underestimatethe changes in net revenue. The IPEDS data do not report the proportion of aid providedto freshmen exclusively. Finally, some of the institutions in the IPEDS database did notreport their amounts of institutional grant aid in both 1989-90 and 1995-96. For completeaccuracy and comparability, institutions with missing data in either year were eliminatedfrom the analysis, thus reducing the number of institutions included in this part of thestudy to 915.
Despite these differences, the patterns and results for the IPEDS census reflect thosereported for institutions in the NACUBO sample. Institutions with above-average increasesin discount rates saw their net revenue decline by $70 per FTE undergraduate (see Table9 on the next page). For each dollar they gained in additional gross tuition and fee revenue.they provided almost $1.03 in tuition discounts. This net revenue loss is lower than theamounts reported for the NACUBO sample, but the difference might be due to theinclusion of tuition revenue from graduate and professional students. Students in advanced
' See 20 USC 1094(417).'Title IV of the Higher Education Act authorizes the Federal Pell Grant program and other programs thatprovide federal financial aid for students to attend postsecondary institutions.
BEST CO
31
27
The patterns andresults for the IPEDScensus reflect thosereported forinstitutions in theNACUBO sample.
AVAII LADLE

degree programs usually are charged higher tuition and feeamounts than undergraduates(National Association of Student Financial Aid Administrators1999), but are less likely toreceive institutional grants. Thus, the inclusion ofrevenue from graduate/professionalstudents might offsetsome of the losses in funds from discounts provided to undergraduates.If the IPEDS data counted the tuition discounts andrevenue figures for undergraduatesexclusively, the net revenue losses might have beeneven larger.
Table 9. Marginal Increase in Institutional Grants, Tuitionand Fee Revenue, and
Net Revenue for Four-Year Private Colleges and Universities,by Level ofChange in Tuition Discount Rates, 1989-90 to 1995-96
MarginalMarginalNet RevenuePercent ofPercentage PointIncase inLactase in(Marginal Rev.MarginalChange in TuitionNumber ofInstitutionalTuition and FeeMinus MarginalRev. Used forDiscount RatesInstitutionsGrants per FITRevenue per FTECost) Pet FITAdded GrantsBelow Avg (1.6 Points or Less)229$162$2,746$2,5845.9%Avg (Between 1.6 to 10.3 Points)4571,4613,1421,68146.5%Above Avg (10.3 Points or More)2292,7612,691-70102.6%Total915$1,477$3,000$1,52349.2%Table 10. Marginal Increase in Institutional Grants,Tuition and Fee Revenue,and Net Revenue for Selective Four-Year Private Colleget andUniversities,by Level of Change in Tuition Discount Rates, 1989-90to 1995-96
MarginalMarginalNet RevenuePercent ofPercentage PointIncrease inIncrease in(Marginal Rev.MarginalChange in TuitionNumber ofInstitutionalTuition and FeeMinus MarginalRev. Used forDiscount RatesInstitutionsGrants per FTERevenue per FITCost) Per FTEAdded GrantsBelow Avg (1.6 Points or Less)35$1,342$6,215$4,87321.6%Avg (Between 1.6 to 10.3 Points)712,4294,5632,13453.2%Above Avg (10.3 Points or More)353,9383,815-123103.2%Total141$2,409$5,110$2,70147.1%Source: U.S. Department of Education 1990a and 1998a; ResearchAssociates of Washington 1999.
28
The decline in net revenue was even greater for selective and highly selectivecollegesand universities with above-average increases in discountrates. Tables 10 and 11 displaythe losses in net tuition revenue for selective and less selective institutions,respectively.Selective colleges and universities with above-average growth in tuition discountrates lost$123 in net revenue per FTE, while less selective institutions with above-averageincreasesin discount rates saw a loss of $39 per FTE undergraduate. Less-than-selectiveinstitutionswith below-average changes in discountrates saw their grant spending per FTE student
BEST COPY AMU&

Table 11. Marginal Increase in Institutional Grants, Tuition and Fee Revenue, andNet Revenue for Less-Than-Selective Four-Year Private Colleges and Universities,by Level of Change in Tuition Discount Rates, 1989-90 to 1995-96
MatginalMarginalNet RevenuePercent ofPercentage PointIncrease inIncrease in(Marginal Rev.MarginalChange in TuitionNumber ofInstitutionalTuition and FeeMinus MarginalRev. Used forDiscount RatesInstitutionsGrants per FTERevenue per FTECost) Per FTEAdded GrantsBelow Avg (1.4 Points or Less)189$ (129)$1,740$1,869NAAvg (Between 1.4 to 10.5 Points)3771,1662,6781,51243.5%Above Avg (10.5 Points or More)1892,4822,443-39101.6%Total755$1,206$2,409$1,20350.1%Source: U.S. Department of Education 1990a and 1998a; Research Associates of Washington 1999. fall by $139, while spending by those with above-average increases in discount rates jumpedby $2,482.
Like the NACUBO data, the IPEDS figures also show that the institutions withabove-average increases in discount rates had the steepest declines in total undergraduateenrollment. Between fall 1989 and fall 1996, institutions with above-average increases indiscounting saw their total number of undergraduates shrink by 5 percent. This declinecompares with a 12 percentgain in the number ofstudents at colleges anduniversities that changedtheir discount rates bybelow-average rates (seeFigure 9). The number offirst-time,full-time,degree-seeking freshmen atabove-average institutionsdeclined by 1percent. Bycomparison, the number offirst-year students rose by3 percent at colleges anduniversities with below-average changes in discountrates.
Institutionsatalldiscount rate change-levelsin the IPEDS database hadsubstantial growth in low-
Figure 9. Percentage Changes in Total Undergraduate Enrollments andEnrollments of First-Time, Full-lime Freshmen at Four-Year PrivateColleges and Universities, 1989-90 to 1996-97 10
12%
5
3%3%2%
2%2%
Below AverageAverage(1.6 Points or Less)(1.6 to 10.3 Points)N=229N=457
-1%
All Institutions
N=915II Total Enrollment
II Freshmen Enrollment
-5%
(10I;N=229BEST COPYAVAILABLe
,),,vientAsvoerrajoere)
Source: U.S. Department of Education 1989 and 1998e.
33
29

income undergraduates. The number of Pell Grant recipientsat colleges and universitieswith above-average gains in discountrates grew by 11 percent (see Figure 10). This growthwas nearly the same as the increases in Pell Grant recipientsat institutions with below-
Figure 10. Percentage Change in Total UndergraduateEnrollments and Pell
Grant Recipients at Four-Year Private Colleges andUniversities,
1989-90 to 1996-97
10
12%
11%10%
Below AverageAverage(1.6 Points or Less)(1.6 to 10.3 Points)N=229N=457
-5%
Above Average(10.3 Points or More)N=229
Sourcc: U.S. Department of Education 1989, 1990c, 19976, and 1998c.
All Institutions
N=915II Total Enrollment
III Pell Grant Recipients
30
average and average changes in discount rates. Both the NACUBO and IPEDS data showthat four-year private colleges and universitieswere able to raise their enrollments of low-income students, but the pricesome institutions paid for this achievementwas lower nettuition revenue and lower total undergraduate enrollments.
34
BEST COPY NAM.

Conclusion t is dear that a large number of institutions used tuition discounting strategies that haveled them down a precarious financial path. In just six years, at least one quarter of thefoul -year private colleges and universities increased their discount rates by 10 percentagepoiits or more, and raised their spending on institutional grants by almost $3,400 perhill-time equivalent undergraduate. These increases occurred at the same time that increasesin gross tuition revenue were smaller, partly because of slower growth in tuition and feeprices. As a result, many institutions lost substantial amounts of net tuition revenue.Even some selective institutions lost more than $800 per FTE undergraduate in netrevenue.
Obviously, losses of this magnitude cannot be sustained for long without affectingspending for other education-related expenditures. The institutions that lost net tuitionrevenue had lower increases in the funds they devoted to instruction and academic supportservices, and actually cut spending on maintenance of campus buildings and grounds.Large losses in net revenue appear to have made it more difficult for these institutions tofund these educational operations.
To make matters worse, the large increases in institutional grants to First-year studentsdo not appear to have significantly affected the undergraduate enrollments at many privatecolleges and universities. In fact, the institutions with the largest increases in discountrates had the smallest increases in numbers of freshmen, and their total undergraduateenrollments fell by 5 percent. Selective colleges with above-average increases in discountrates saw their total enrollments shrink by 8 percent.
Increasing institutional grants also does not appear to have influenced substantiallythe academic characteristics of the undergraduate student bodies at most campuses. Despitelarge increases in the proportion of undergraduates who received merit scholarships andother non-need-based grants, even the most selective institutions saw the median SATscores of their first-year students rise by less than I0 percent during the study period.These small increases suggest that the institutions' efforts to provide large non-need-based aid to students from upper-income families thus far have not substantially helpedinstitutions raise their academic profiles or rankings in college guidebooks. Median SATscores at less selective-institutions actually fell by about 2 percent. These institutions mayhave been more concerned with raising their total enrollments than with raising theacademic profiles of their entering classes.
These negative trends may have occurred because, at many institutions, enrollmentmanagement goals have focused on increasing freshmen enrollment without alsoemphasizingretentionof all undergraduates toward degree attainment. The colleges anduniversities with the highest increases in discount rates also had the lowest six-yeargraduation rates. Student retention might be a bigger problem for these institutions. Ifthese colleges and universities had devoted more resources to services and programs thathelp retain studentsand had slightly lower increases in tuition discountingtheymight have increased enrollments, net revenue, and graduation rates.
However, tuition discounting does appear to have helped to increase the numbers o6low-income undergraduates at four-year private colleges and universities. The number ofPell Grant recipients grew substantially at all selectivity levels, despite increases in tuitionE
31
Losses of thismagnitude cannot besustained for longwithout affectingspending for othereducation-relatedexpenditures. prices. This strongly suggests thata large share of the institutional grant dollars are stillbeing directedto low-income undergraduates. These funds appear to be providingeducational opportunities foran increasing number of undergraduates from low- andmoderate-income families.
Increasing tuition discounts appears to have made it possible formore undergraduatesto enter higher education. Unfortunately, a number of colleges and universities have paida steep price to reach this laudable goal. The rapid increases in discounting have resultedin losses in net revenue, have not improved retentionor graduation rates, and have causedinstitutions to decrease spendingon instruction and other vital services to students. Privatecolleges and universities will haveto try other strategies before even more institutionsfollow a path heading toward disaster.