Expanding Student Options in Higher Education · THE POLICY OF CHOICE: Expanding Student Options in Higher Education viii In 1999-2000, 62 percent of all first-year under-graduates

The Institute for Higher Education Policy

Prepared for The New Millennium Project onHigher Education Costs, Pricing, and Productivity

S P O N S O R E D B Y :

The Institute for Higher Education PolicyThe Ford Foundation

The Education Resources Institute

the policy of choiceExpanding Student Optionsin Higher Education

Prepared for The New Millennium Project onHigher Education Costs, Pricing, and Productivity

S P O N S O R E D B Y :

The Institute for Higher Education PolicyThe Ford Foundation

The Education Resources Institute

the policy of choiceExpanding Student Optionsin Higher Education

August 2002

The Institute for Higher Education Policy is a non-profit, non-partisan organization whose missionis to foster access to and quality in postsecondary education. The Institute’s activities are designed topromote innovative solutions to the important and complex issues facing higher education. These activitiesinclude research and policy analysis, policy formulation, program evaluation, strategic planning andimplementation, and seminars and colloquia.

For further information, please contact:The Institute for Higher Education Policy1320 19th Street, NW, Suite 400, Washington, DC 20036Phone: 202-861-8223/Facsimile: 202-861-9307/Internet: www.ihep.org

iii

contents

Preface ............................................................................................................................................. v

Executive Summary ....................................................................................................................... vii

Introduction .................................................................................................................................... xi

The Concept of Choice .................................................................................................................... 1

The College Choice Process ............................................................................................................ 9

Outcomes: The Status of Choice in a Policy Context ........................................................................................... 13

A Level Deeper: Examination of Financial Aid Patterns ....................................................................................... 29

Conclusions .................................................................................................................................. 39

References ..................................................................................................................................... 41

iv

v

T his report is the latest in a series of reportsand papers published under the aegis of theInstitute for Higher Education Policy's New

Millennium Project on Higher Education Costs,Pricing, and Productivity. Sponsored by the Institutefor Higher Education Policy, the Ford Foundation,and The Education Resources Institute (TERI), theProject is a multi-year effort to improve under-standing and facilitate reform of the complexsystem for financing higher education.

Given the upcoming reauthorization of the HigherEducation Act, the goals and effectiveness of federalTitle IV student aid programs once again have cometo the forefront of higher education policy debate.Particularly important is the role of federal studentaid in promoting access, choice, and affordabilitydespite rising college tuitions and increasing numbersof disadvantaged students who desire to enroll inpostsecondary education. Within this context, wefelt it was important to reexamine the issue ofcollege choice, in part to lay the groundwork for adiscussion of the appropriate goals for federal

student aid and the effectiveness of federal aidprograms in meeting those goals. This report buildsupon a prior report in this series, State of Diffusion,which outlined the various purposes of student aid.Other reports and papers published under the NewMillennium Project have examined such issues as thepublic and private benefits of higher education,changes in tuition policies, higher education costmeasurement, and the effects of federal tax credits.

The report was authored by Alisa FedericoCunningham, Director of Research, who would like tothank her colleagues at the Institute. MelissaClinedinst, Jamie Merisotis, Christina Redmond,Jessica Shedd, Jane Wellman, and Tom Wolaninprovided helpful comments, ideas, and substantivefeedback at various stages in the writing of this report.In addition, Loretta Hardge and Deanna High contrib-uted greatly to the editing, framing, and production ofthe report. The author acknowledges the contribu-tions of these individuals to this report and recognizesthat they are not responsible for any errors ofomission or interpretation contained herein.

preface

vii

P ublic policy supporting higher educationis largely framed around the idea ofaccess—students with the desire and ability

to go to college should be able to attend some typeof postsecondary institution. While this focus onaccess to higher education persists at the federaland state levels, there is increasing interest in animportant and related issue: access to what?

Historically, the ability of qualified, motivated stu-dents to enroll in some type of postsecondaryinstitution and to pick a school has received varyinglevels of policymaker support. Discussions aboutcollege choice are increasingly common, even if theterm “choice” is not used. For example, the appar-ent inability of financial aid (especially grants) tomatch the pace of rising tuition is central to discus-sions about student choice, since financial factorsroutinely constrain student choices. Similarly,debates about affirmative admissions in highereducation are intertwined with the idea that aqualified student should be able to pick the collegeor university of his or her choice without race beinga fundamental determinant of admission or denial.Further adding to the interest in choice at a policylevel are reports that prominent new approaches tocollege financing—such as the federal HOPE taxcredits, or the growing number of state merit-basedfinancial aid programs—appear to have moreinfluence on where students go to college than onwhether they go at all.

Compounding policy discussion about collegechoice is that the word “choice” has differentmeanings in the policy context. At the K-12 level,for example, “choice” has come to mean the ability

executive summaryof parents to send their children to a school thatthey choose, in many cases through the use ofvouchers to reduce the price differential betweenprivate, tuition-based schools and free publicschools. However, in higher education, whichuniversally relies on tuition whether the institutionis public or private, promoting college choice hasbeen a widely supported goal of federal, need-based financial aid programs (such as Pell Grants)since at least the 1970s. The goal of choice isexplicitly encouraged by inclusion of the price ofcollege in the federal need analysis formula, inwhich the level of student “need” increases if astudent chooses an institution with a higher price ofattendance, all else being equal.

Given the ongoing interest in college choice, thisreport takes a fresh look and revisits what collegechoice means and how it works. Exploring in somedetail the various definitions used to describechoice, the report reveals what is known about thefactors that influence students’ decisions, and thecurrent “status” of choice. Throughout the report,the analyses focus on choice in relation to federal,need-based financial aid policy.

Key FindingsHigher education policy discussion about choicehave evolved from focusing on whether studentshave a choice of public versus private institutions toexamining whether students can choose between atwo-year college, a for-profit institution, or a four-year institution. The following key facts representwhat is known about the current status of choice inU.S. postsecondary education:

THE POLICY OF CHOICE: Expanding Student Options in Higher Education

viii

� In 1999-2000, 62 percent of all first-year under-graduates were enrolled at public two-yearinstitutions; 19 percent attended public four-year institutions, 9 percent attended privatenot-for-profit four-year institutions, 8 percentattended proprietary schools, and the rest (2percent) attended other types of institutions.

� First-year undergraduates with non-traditionalcharacteristics were more likely to attendlower-priced, two-year institutions and propri-etary institutions than were traditionalfreshmen. For example, 75 percent of students24 years or older were enrolled at public two-year institutions, compared to 41 percent ofstudents 18 years or younger.

� The lowest-income dependent students weremore likely to attend public two-year institu-tions (55 percent) and proprietary institutions(8 percent) than the highest-income dependentstudents (45 and 2 percent, respectively).

� Black and Hispanic students were more likelyto attend proprietary schools than werestudents of other races/ethnicities, 12 percentfor Black and Hispanic students compared to6 percent for white students.

The data reveal that patterns of enrollment havechanged over the period 1989-90 to 1999-2000,suggesting a movement toward lower-pricedinstitutions (especially community colleges) formany groups of students. The trends break downdifferently for different types of freshmen.

� Among traditional freshmen, the proportion oflow-income freshmen decreased at most types offour-year institutions, but increased at public two-year institutions. On the other hand,higher-income students decreased at less selec-tive four-year schools and proprietary institutions,but increased at both public two-year institutionsand more selective, PhD four-year institutions.

� Among non-traditional freshmen, the propor-tion enrolled at proprietary schools decreased,

but increased at public two-year institutions—especially among students from the lowerincome categories.

The report also examines whether aid is beingtargeted in a way that would provide choice, giventhe framework of federal need analysis. Thisapproach assumes that if aid is promoting choice,students attending more expensive institutionswould be more likely to receive financial aid, and toreceive higher average amounts, within the samelevel of family income. This approach also assumesthat financial aid is helping to promote choice if thecosts of choosing a more expensive institution arereasonable. In fact:

� At more expensive institutions, higher propor-tions of full-time, full-year dependentundergraduates received aid from any source,across all family income categories.

� The percentages of students receiving institutionalaid and loan aid were tied more to the type ofinstitution than to family income, suggesting thatthese types of financial aid were promoting choicemore than federal or state aid, or grants.

� The choice of a more expensive institution isrelatively more costly for students from low-income backgrounds than for students withhigh-income backgrounds.

� In 1999-2000, a substantial proportion of federalneed-based financial aid (from 40 percent tomore than 70 percent) for all students, depend-ing on definition of choice, went towardpromoting choice. However, the proportionvaried even more by family income, withincreasing percentages of aid going to promotechoice in the higher income categories.

ConclusionsThis report suggests that there are several keyquestions to consider in future policy discussions ofcollege choice:


ix

� Why are certain groups of students lesslikely to attend specific categories ofinstitutions? Taken altogether, the findingsreveal declining choice for certain types ofstudents, apparently confirming perceptionsthat many students are choosing lower-pricedinstitutions due to increasing tuition levels.However, enrollment patterns alone are notenough to definitively conclude that disadvan-taged students’ barriers to choice are solelymonetary. A longitudinal perspective involvingthe entire choice process suggests that thereality is complex, and that monetary factorsare important but not alone in influencingeducational opportunity.

� How should choice be defined for publicpolicy purposes? The specific meaning ofcollege choice depends on one’s perspective. Insome cases, “choice” is a codeword for specificpolicy goals; in this sense, choice may mean notjust the enabling of options for students, butalso may connote ideas about institutionalquality and equalizing tuition levels acrosssectors. Sharpening the definition of choice andclarifying the role of specific forms of financialaid in promoting choice will enable bettertargeted public policy.

� Is federal need-based aid currently awardedin a way that reflects the goal of promotingchoice? The findings present a mixed picture.Conclusions depend upon the definition of choiceand the income background of the student.

� Should choice be a goal of federal aidpolicy? The existing framework defines financial“need” in a way that is sensitive to the price ofattendance, in order to support the goal ofchoice. However, this means that students canbe defined as “needy” based upon the price ofthe institution they choose, rather than as aresult of disadvantaged economic circumstances.What is the appropriate balance?

� What is the likely consequence of offeringmore, or less, choice to students? Not onlydo public policies change over time, but alsothe composition of students attendingpostsecondary education institutions and thestructure of higher education itself shift inresponse to competitive pressures. Given theongoing patterns of tuition increases—withgrowing gaps between two-year and four-yeartuitions, and public and private tuitions—choicemay have eroded even without changes infederal policy. This deserves further examina-tion and analysis.

xi

O ver the past few decades, collegeenrollment has been increasing despiterising tuition levels and the apparent

inability of financial aid (especially grants) to matchthe pace of tuition. Observers have tried to under-stand the incongruity between increasingenrollment and increasing net prices by focusing onthe issue of where students are enrolling. But thepostsecondary education system has become verydifferentiated—ranging from open admissions, two-year schools to highly selective, restrictedadmissions research universities—and the differ-ences are reflected in wide variations in tuition andenrollment patterns. Low-income students, forexample, appear to be more sensitive than middle-or upper-income students to college prices (Heller1997), and disadvantaged students are less likely toenroll in selective private institutions or even four-year institutions. Thus, it may be possible that thecollege financing system is effective in making someform of postsecondary education widely accessible,but is less successful in matching access to a suitablepostsecondary experience with the abilities anddesires of some groups of students.

In practice, the process of choosing where toattend college is highly complicated, and issues oftuition and financial aid are not the only ones thatinfluence students’ decisions. College choices alsoare constrained by non-monetary factors such assubject of study, institutional admissions practices,family or work responsibilities, and family/societalexpectations (Choy and Ottinger 1998). In fact,college enrollment patterns depend on self-selec-tion processes that take into account all of thesefactors (McDonough 1997), which narrow the

introductionrange of institutions that a student considers. Thisprocess must be kept in mind when enrollmentpatterns are used to assess how effectively thecollege financing system provides choice. Never-theless, these patterns (and how they have changedover time) are worthy of examination since thenature of an institution attended—especially thelevel of the institution—affects the extent of thestudent’s education as well as his or her eventualsocial and economic return (Hearn 1991;McDonough 1997).

The issue of college choice continues to be raised inhigher education policy circles in a variety of con-texts. In recent years, for example, someresearchers have asserted that low-income studentsare increasingly becoming concentrated in lower-priced, two-year institutions, reflecting a growingstratification of postsecondary education by socio-economic status (McPherson and Schapiro 1998).Others have suggested that new student aid initia-tives, such as the federal HOPE tax credits, haveserved to encourage choice rather than access topostsecondary education (Wolanin 2001). Given thecontinuing salience of the issue, it makes sense toreexamine the concept of college choice within thecontext of public policy.

This report takes a fresh look at college choicefrom several perspectives—the various definitionsused to describe choice, what is known about thefactors that influence students’ decisions, and thecurrent “status” of choice. The report is dividedinto corresponding sections. The first sectionpresents different definitions of “choice” as theterm has come to be used in the policy context,


xii

and explains how the definitions have expandedor changed over time. This section lays out aframework for understanding college choice in apolicy context. Next, the report explores thechoice process and the factors that influence eachstep of the process, highlighting some of thesubstantial research that has been conducted inthe area. In the last section, the report focuses onthe outcomes of college choice decisions byreexamining previous research using more recentdata. Data are drawn primarily from the NationalPostsecondary Student Aid Study (NPSAS), a U.S.Department of Education dataset that capturesinformation on undergraduate enrollment,

demographic characteristics, financial aid patterns,and other characteristics at one point in time.This section highlights the current status of choiceas reflected in the distribution of students byinstitutional type.

Throughout the report, the analyses focus onchoice in relation to federal student aid policy,while acknowledging the importance of otherpolicy instruments such as state financial aid andearly intervention programs. The report offers aframework for future conversations about collegechoice and the role of federal student aid inpromoting choice.

1

T he concept of choice has been defined indifferent ways, depending on the stake-holder’s perspective—students, colleges and

universities, federal and state governments, andothers. Within the public policy context, the defini-tion of choice has specific ramifications when seen asa goal of federal student aid policy. In addition, policydebates regarding college choice tend to shift overtime, reflecting new concerns and goals.

Defining choiceThe concept of educational “choice” has been usedin many contexts, from elementary and secondaryeducation to undergraduate education and beyond.At the K-12 level, public policy debates aboutchoice tend to revolve around proposals of publicfunding for school voucher programs that enableparents to enroll their children in private or paro-chial schools or, more recently, the choice ofalternatives within the public school framework,such as charter schools and magnet schools. Publicpolicy debates about choice at the postsecondaryeducation level differ, stemming largely from thefundamental distinction that K-12 education iscompulsory, whereas all of postsecondary educa-tion represents “choice” to some extent.

At the broadest, most abstract level, the concept ofcollege choice involves a student’s ability to attendthe college or university most suited to his or herdesires and academic abilities. Researchers routinelydefine “choice” as the end result of a process, in

the concept of choicewhich students evaluate their options and chooseamong them. In fact, significant theoretical ground-work exists that describes the college choice processfrom the perspective of an individual student, usuallyin terms of multiple stages (see, for example, Manskiand Wise 1983). Briefly, a student first decideswhether to attend college. In the second phase—search and application—students seek informationabout colleges and formulate a choice set, a range ofcolleges to which they apply. In the final selectionphase, students evaluate their choice sets (dependingon where they are admitted) and make selectiondecisions. (The choice process is described in moredetail in the following section.)

For the purposes of government policymaking,choice is frequently defined in relation toaccess. Both are aspects of educational opportunity.In the abstract, “access” can be defined as the abilityto attend college; providing access to a studentmeans that he can attend college. Access can bemore specifically defined in a number of ways, fromthe ability to enroll in any form of postsecondaryeducation at any time, to the ability to enroll in afour-year college immediately after high school (seeInstitute for Higher Education Policy 1999b). Choiceraises the question: access to what? Thus, providingchoice to a student goes beyond access, to meanthat a student can attend an institution appropriatefor his or her desires and academic abilities. Whatthat means specifically depends on which definitionof access is at work, and in fact, the two concepts—access and choice—are integrally linked.1

1 Another perspective related to the definition of choice is that of colleges and universities, where choice may be perceived in light of institutions’ desires for diverse studentbodies, or within the context of increased marketing efforts. Much of the financial aid awarded to students by higher education institutions is offered to encourage a specific“mix” of students, or to allow students from specific backgrounds to choose to attend a particular institution.


2

In order to judge the effectiveness of governmentpolicy in encouraging choice, the concept needs tobe operationally defined, within the public policycontext. Specifically, college choice often refers tooutcomes—i.e., the pattern of enrollment atvarious types of institutions. Although suchoutcomes may, at first glance, appear to be afunction of the final stage of the student choiceprocess (selection), they are actually a cumulativeresult of the process and are influenced at manypoints along the path. This influence is importantto keep in mind when using outcomes data todraw conclusions about specific policies.

In practical terms, the focus on outcomes variesbased on the specific groups of institutions exam-ined. Examining choice through outcomes dependslargely on the definition of the “institution of basicaccess,” or the type of institution that virtuallyguarantees access to postsecondary education. Forexample, some argue that public two-year institu-tions, because they receive substantial (mostlystate) subsidies to keep tuition low and provide awide range of academic and vocational programs,should be the schools of basic access; others arguefor public four-year institutions, sincepostsecondary education should be primarilyacademic and a postsecondary education shouldresult in a four-year degree. In examining out-comes, institutions of choice would be the schoolsthat are somehow higher/better/more expensivethan the institutions of basic access. Distinguishingbetween access and choice implies that providingaccess to the lowest-priced institution does notnecessarily match students with institutions thatbest suit their needs and abilities.

Choice and federal aid policyVarious types of public policies aim to encouragecollege choice. These policies rely on a wide rangeof approaches from early intervention programsand early awareness activities, to tax exemptionsand benefits to private institutions. In addition,promoting college choice is a specific goal of federalfinancial aid, especially need-based aid programs.Since the 1970s, one goal of federal need-based aidhas been to provide choice for the individualstudent recipient.2 The price of college was there-fore included in the federal need analysis formula,which is used to determine the expected familycontribution (EFC) and the amount of federal need-based aid a student may receive, so that aidprograms could address choice as well as access.

Both access and choice are explicit goals of federalneed-based aid policies. Need-based student aidattempts to promote access in general by reducingstudents’ net price of attendance. Aid attempts tofoster choice in particular by taking into accountthe variation in price of attendance at differenttypes of postsecondary institutions. This is accom-plished primarily through a feature of the federalneed analysis formula, in which the level of student“need” increases as a student' s price of attendanceincreases, all other factors equal.3

To an extent, access and choice have become codewords in controversies over federal student aidprograms, related to who receives aid, what type ofaid is received by certain students, and whichinstitutions are involved. Of course, access andchoice cannot be completely separated, as aid may

2 This goal is intertwined with the decision to use student aid as a portable mechanism for the distribution of resources, rather than providing federal dollars directly toinstitutions.

3 Note that these definitions of access and choice are distinct from the concept of affordability. Affordability refers to whether the money students and their families actuallypay to attend college is within their reach (Institute for Higher Education Policy 1999). When analyzing how financial aid reduces the total price of attendance faced bystudents, it is defined differently based upon the goal of aid examined; thus, access (and choice) may be analyzed by looking at price less all financial aid received, whileprice minus all grants is a better measure for affordability (Cost Commission 1998). This is based on the premise that loans (and work-study) are equal to grants inmeeting immediate financial need, and therefore enable access in the short-term; in the long run, however, loans do not reduce the net price to the student because theymust be repaid (or, in the case of work-study, must be earned through work).


3

affect both simultaneously—it may change enroll-ment patterns and encourage students to enroll whootherwise would not attend college. Nonetheless, atrade-off exists between using a limited amount ofneed-based aid primarily to ensure access (throughemphasizing a student’s economic background) or topromote choice (by allowing price of attendance toinfluence the amount of a student’s need). Therelative emphasis on access and choice can bealtered by changing aspects of the federal needanalysis methodology (CBO 1991).4

From the perspective of federal need-basedstudent aid policy, more expensive institutionsbecome the “schools of choice.” Thus, analysesof outcomes tend to focus on enrollment ofdifferent groups of students at more expensiveschools versus less expensive schools. Opera-tionally, outcomes may refer to attendance atprivate institutions over public institutions,5 four-year institutions over two-year institutions,selective admissions institutions over non-selective, open admission schools, andhigh-spending schools over lower spendinginstitutions, among other definitions.

Of course, as mentioned above, choice of collegedepends on many factors, and a student may notnecessarily prefer a more expensive school to alower-priced institution. Within this context, federalstudent aid aims to enable students to choose moreexpensive institutions if they are accepted by thoseinstitutions and desire to attend. In examining out-comes, then, ideally the effect of financial aid onchoice ought to be isolated from the effects of otherfactors. Yet in reality, academic and other non-monetary factors are too closely intertwined withmonetary factors to disentangle completely.

Policy changes over timeThe context of public policy relative to collegechoice has changed over time. Although this sectionfocuses specifically on federal financial aid policy, itis important to keep in mind that simultaneouschanges have occurred in state policy, as well,focusing on other factors in the choice process. Forexample, increasing federal and state-level atten-tion is being directed to early intervention efforts toequip students to choose. In addition, the optionsthat influence student choice are changing, such asadvances in technology related to distance learning,making proximity less relevant for a significantproportion of the student population.

In the 1970s, commitment to choice as a goal offederal aid was made explicit in statements thatfederal need-based grants would “not only guaran-tee low-income students access to publicinstitutions, but also provide a modest level of choicebetween public and private institutions” (ACSFA2001, 2). This reflected the reality that in those earlyyears, choice was conceived in terms of the ability toattend private, high-priced institutions. Theseinstitutions were perceived to offer substantialadvantages, such as personalized attention andbetter success in graduating students, and to contrib-ute to diversity within the higher education sector.6

At that time, there was concern about the growingtuition gap between the public and private sectors,suggestions of future financial difficulties, and fearsthat the private sector was in jeopardy, as its propor-tion of enrollment continued to decrease (Brenemanand Finn 1978).

Reports such as that by the Carnegie Commissionon Higher Education, which advocated raising

4 For example, the proportion of the total price of attendance that is counted can be changed.5 The context of private versus public institutions was originally conceptualized in terms of four-year, not-for-profit institutions. However, the growth of the private for-profit

sector has blurred the public/private distinction within the choice framework. It is useful to keep in mind that private for-profit institutions tend to be more expensive thanpublic institutions, especially public two-year institutions; thus, in some analyses, private for-profit institutions are considered along with private not-for-profit institutionsdue to their higher average prices. In other analyses, private for-profit institutions are grouped with public two-year institutions because both sets of institutions tend tooffer certificates and associate’s degrees rather than bachelor’s degrees or higher.

6 As mentioned in the previous footnote, “private” institutions generally meant private not-for-profit institutions.


4

public tuition to narrow the gap, were accompa-nied by political battles that deterred efforts toaddress the issue.7 As a result, federal governmentattempts to help the private sector were piece-meal and generally came in the form ofneed-based, student aid strategies to encourageor equalize choice (Gladieux and Wolanin 1978;Wolanin 1998).8 Several key provisions wereenacted in 1972:

� The Basic Equal Opportunity Grant (BEOG, nowPell Grant) half-cost rule, in which grants couldnot exceed one-half the cost of attendance;

� The campus-based funding trigger, whichrequired campus-based programs to be fundedbefore the basic grants could become opera-tional; and

� The State Student Incentive Grant (SSIG, nowLEAP) program, which encouraged the devel-opment of state scholarship programs and hadspecific provisions encouraging scholarships forstudents in the private sector as well as inpublic institutions.

During the 1980s, the private sector fiercelyfought to keep these provisions despite a generalretrenchment in federal student aid policy(Wolanin 1998). Meanwhile, this decade also sawthe beginning of a shift in federal student aid policytoward loans instead of grants. Combined withwidening tuition gaps and the nature of the needanalysis formula—in which one student' s definedneed may surpass that of another solely due toattendance at a higher-priced institution ratherthan a lower ability to pay—this shift has hadimplications for the ability of different forms offinancial aid to encourage choice. For low-income

students, need-based federal grant aid such as PellGrants became less likely to encourage choice, asthe grants covered a declining proportion of theaverage price of attending a four-year institution(especially selective and/or private institutions),and therefore were unlikely to encourage enroll-ment shifts. On the other hand, need-based aid tohigher-income students, especially through loanprograms, was more likely to subsidize choicerather than access.

By the early 1990s, providing choice was no longera specific mission of the Pell Grant program. Inaddition, the campus-based trigger was repealed in1986, and the half-cost provision was repealed in1992 (Wolanin 1998). Choice has remained a goalof the larger student aid system, nonetheless, andthe federal need analysis formula continues toinclude price of attendance as one of its elements.As price differentials widen between types ofinstitutions, for example, increasing numbers ofstudents have turned to federal student loans tohelp pay for higher tuitions and living expenses.Although the federal tax credits enacted in 1997are still too new for a comprehensive evaluation oftheir effects, observers believe that studentsattending institutions with higher prices of atten-dance benefit most from the credits (Wolanin2001); the credits therefore may be enablingmiddle-income students to continue to enroll inmore expensive institutions than they otherwisemight be able to afford, or at the very least may beinfluencing the perception of institutional alterna-tives. In a sense, the goal of encouraging collegechoice has been left to other forms of financial aid.

In parallel with the shift in financial aid goals,changes in the implicit definition of choice used by

7 The Carnegie Commission on Higher Education (1973) suggested narrowing the tuition gap by increasing public tuition levels a bit, while offsetting the higher tuition withincreased availability of aid for lower-income students.

8 Federal involvement with private institutions also came in the form of direct payments to institutions (for example, for research and development contracts) and taxprovisions regarding donations to non-profit institutions (Breneman and Finn 1978). State policy has been more explicit in its support of private colleges, and state studentaid programs are generally open to students attending public or private institutions (Gladieux and Wolanin 1978). In fact, the Carnegie Commission on Higher Education(1973) suggested that state governments take into account the impact of their decisions on the private sector, and recommended the use of SSIG/LEAP incentives toincrease state aid to students in both the public and private sectors, while addressing the tuition gap.


5

policymakers were also occurring. In the early1970s, choice tended to be conceptualized interms of the ability to attend private, mostlyselective institutions. For example, the CarnegieCommission on Higher Education (1973) framedchoice in terms of private institutions, which itstated had the same social and economic benefitsas public institutions—thereby offering a justifica-tion for public dollars to have the goal of choice.In more recent years, the term “choice” has beenincreasingly applied to the ability of disadvantagedstudents to attend any four-year institution, publicor private.9 In part, this change has stemmed froma growing concern that low-income students arebeing concentrated in two-year institutions. It alsohas been influenced by data consistently showinghigher earnings for bachelor’s degree recipientsthan for those with some college (Carnevale andRose 1998) and by recent research suggesting thatthe personal benefits (in terms of future income)of attending highly selective institutions haveincreased over time (Hoxby 1998). Ultimately, thisshift in definition may signal a change in thetheoretical basis of choice policy.

A framework for understanding choiceClearly, “college choice” can mean different thingsto different people, and these meanings havechanged over time. To evaluate public policy, it isimportant to have measurable indicators of effec-tiveness; yet in the case of “choice,” outcomemeasures result from a cumulative, complicatedprocess. In thinking about public policy and collegechoice, it is helpful to keep in mind the multipleaspects of the concept. Toward this end, FigureOne summarizes a framework for understandingthe concept of choice, from various perspectives.

From the student perspective, several types offactors influence the choice process:

� Institutional characteristics, such as the locationof the school, the spaces available, and theprograms offered;

� Academic factors, including a student’s level ofacademic preparation as well as the admissionscriteria of the institution; and

� Monetary factors, such as the tuition and feelevels, financial aid offered, and the students’and families’ ability to pay.10

These factors influence a student’s choice atvarious stages in the process, from the develop-ment of aspirations to go to college, to theconstruction of a “choice set” (a group of institu-tions to which the student applies), to the selectionof the institution at which the student will enroll.

The public policy perspective, vested in society’snotion that some degree of choice is desirable, aimsto encourage student choice. The options for publicpolicy range along a continuum of degrees of choice:

� At one end is the “legal” definition of choice, inwhich students are not by law excluded fromchoosing any institution, although in practicethey are constrained by the three groups offactors above.

� At the other extreme is the definition of “fullchoice,” in which all institutions would be opento all students equally. The three groups offactors would no longer restrict students’choices or perceptions of available choice sets.

The first definition is too limited; the second, tooexpansive. Public policies generally aim at a “bal-ance” somewhere in the middle of the continuum.Specifically, federal need-based student aid policy(through need analysis) attempts to neutralize orreduce the influence of monetary factors on astudent’s choice decision. Put another way, if a

9 Interestingly, there has been a similar shift in public policy debate about choice in elementary and secondary education, in which the debate about public funding for privateschool attendance has been recast to an extent to focus on choice among public school alternatives

10 These factors are not truly separate from each other; in particular, a student’s socioeconomic background is integrally linked with academic preparation (see, for example,Berkner and Chavez 1997).


6

(pric

es, a

id, a

bilit

y to

pay

)

acad

emic

fac

tors

inst

itut

iona

l ch

arac

teri

stic

sm

onet

ary

fact

ors

Cho

ice

Pol

icy

Con

tinu

um

Fina

ncia

l Aid

Cho

ice

Pol

icy

Con

tinu

um

(pre

para

tion,

adm

issio

nscr

iteria

)(lo

catio

n, a

vaila

ble

spac

es, p

rogr

ams)

"Leg

al"

defin

ition

of

choi

ce =

abl

e to

choo

se, b

utco

nstr

aine

d by

mon

etar

y,

acad

emic

, and

inst

itutio

nal

fact

ors

Full

choi

ce =

all i

nstit

utio

nsop

en to

all

stud

ents

, with

out

cons

trai

ntG

oal o

f pol

icy

= to

ach

ieve

a le

vel o

f cho

ice

som

ewhe

re in

bet

wee

n

Goa

l of f

inan

cial

aid

pol

icy

= to

red

uce

the

effe

ct o

f mon

etar

yfa

ctor

s in

cho

ice,

ass

umin

g ot

her

fact

ors

equa

l

Abi

lity

to a

tten

d a

mor

e ex

pens

ive

inst

itutio

n th

an th

e le

ast e

xpen

sive

one

Abi

lity

to a

tten

d an

yin

stitu

tion,

eve

n th

em

ost e

xpen

sive

one

2-ye

ar v

s.

4-ye

arpu

blic

vs.

priv

ate

nons

elec

tive

vs. s

elec

tive

Figu

re O

ne


7

student is academically qualified for a particularinstitution and the institution has the necessarycharacteristics desired, then the student should beable to attend that institution; monetary factorsshould not be a barrier to that choice.

In practice, federal need-based aid policy generallytries to enhance choice without necessarily equaliz-ing it. The issue becomes what amount of pricedifference between institutional alternatives isreasonable to allow the student some level ofchoice. Federal aid policy regarding choice can beviewed as ranging along a sub-continuum, whichaddresses the question of “what kind of choicedoes financial aid enable?” The various goals forchoice are integrally linked to definitions of access,another goal of need-based financial aid, and theassumption is that non-monetary factors (academicand institutional) are met.

� Assuming access is defined as the ability toattend a (any) postsecondary institution, thenthe lowest-priced institution would be theinstitution of basic access from a need-basedfinancial aid perspective. Therefore, choicewould be defined as the ability to attend anyinstitution more expensive than the least expen-sive one. This definition is the most limited,requiring the least amount of financial aid.

� If access is defined as the ability to attend a(any) four-year institution, then the lowest-priced four-year institution would be theinstitution of basic access from a need-basedfinancial aid perspective. Therefore, choicewould be defined as the ability to attend anyfour-year institution more expensive than theleast expensive four-year institution.

� At the other extreme, access might be definedas the ability to attend a (four-year) institution“matched” to a student’s abilities, in terms ofquality and selectivity. In this case, the institu-tion of basic access would be the lowest-priced(four-year) institution that admits a particularstudent. And, choice might be defined as the

ability to attend any institution that admits thestudent, even the most expensive (mostselective) one. This definition would requirethe most financial aid.

In the abstract, one can attempt to measure theeffects of financial aid on choice by looking at thedistribution of enrollment, given the actual prices ofdifferent types of institutions and controlling fornon-monetary factors. Holding academic andinstitutional factors equal, financial aid would beexpected to somewhat equalize the distribution ofstudents from different income backgrounds byinstitutional type. For more limited definitions ofchoice, this might mean examining two-year versusfour-year institutions; for the most expansivedefinitions of choice, this might mean looking atnon-selective versus highly selective institutions.

In practice, however, outcome measures frequentlycannot fully control non-monetary factors (espe-cially given that academic factors are integrallylinked to monetary factors throughout the choiceprocess). Although the outcomes data may suggestthat choice has not been fully realized (the “status”of choice), the data do not clearly establishwhether the “cause” is due to financial aid, or non-monetary factors, or some combination of both.The timing of causal influences, i.e., the point in thechoice process at which factors have an impact,also cannot be determined from outcomes data.

Longitudinal analyses nonetheless can be used toexplore the reasons why students from certainbackgrounds may not be enrolling at certain groupsof institutions at the same rates as other students.Understanding how students make enrollmentdecisions (through the choice process) can helppolicymakers decide how and where to interveneto improve progress toward the goal of “choice,”however defined. In combination with outcomemeasures, the analysis can point to progress as wellas to areas of concern. The remainder of thisreport surveys both types of research on collegechoice: first, a summary of the choice process

.


8

literature and the factors influencing the process;second, a look at updated enrollment outcomes,using several analytic frameworks.

Choice is not a static concept. Definitions andmeasures of choice change over time, and areaffected by broader changes within thepostsecondary education industry and the economyas a whole. For example, the framework presentedabove does not fully take into account distancelearning as an emerging dimension ofpostsecondary education, yet such technologieshave certainly had an impact on the educationalopportunities and preferences of students. In a

related issue, the private, for-profit sector in all ofits various forms does not fit neatly into this frame-work, especially regarding outcome measures, yetthe growth of the sector makes it important toinclude in any analysis. Finally, much of federalchoice policy was formulated in a time when“traditional” students—those enrolling straightfrom high school, attending on a full-time basis, anddepending on their parents for finances—made upthe majority of college students. Today, the over-whelming majority of students have at least onenon-traditional characteristic (NCES 2002b), andenrolled students reflect a diversity of goals,interests, and preferences.

9

O ver the years, substantial research hasbeen conducted in the area of collegechoice from the perspective of students,

especially regarding the factors influencing studentenrollment behavior.11 In general, this body ofresearch is consistent on the choice process andthe factors that influence each stage.

Stages of the choice processMany researchers have described choosing acollege or university as a multiple-stage process.The process can be a lengthy one—even extendingback to the earliest influences on college aspiration.Each student then passes through a series of stagesin which he narrows his options to a single set ofinstitutions before final selection. The process alsocan be described as a sequence of decisions affect-ing postsecondary education choices: application tocollege, admission, financial aid offers, the choice ofschool versus work, and persistence in college. Infocusing on the choice to enroll at a specificcollege, a student’s institutional alternatives fre-quently are predetermined—a result of priorstudent applications and college admissiondecisions. In fact, individual application decisionsmay be more important than college admissiondecisions in determining attendance, whichsuggests that policymakers must try to influencethe process at an early stage in order to effectchange (Manski and Wise 1983).

the college choice processFor this report, a three-stage model of the choiceprocess is most useful. The model by Hossler andGallagher (1987) describes the stages as: 1) predis-position—reviewing postsecondary opportunitiesand forming college aspirations; 2) search—search-ing for information about possible institutions andformulating a choice set, and group of institutions towhich they apply; and 3) choice—selecting aninstitution from among the choice set to attend.Each of these stages is associated with a specific agegroup—for example, predisposition in grades 7 to 9,search in grades 10 to 12, and choice in grades 11 to12—yet the three stages interact with each other incomplex ways (Cabrera and La Nasa 2000).

Together, studies of the choice process have foundthat different kinds of factors—student and institu-tional, academic and nonacademic—influence thechoice process at various steps along the way.

� Predisposition (aspiration). In this early stage,student background characteristics (income,ability, parent attitudes) and school activitiesaffect a student's decision about whether tocontinue on toward college (Hossler andGallagher 1987). Parental encouragement hasemerged as the most influential variable, andintervenes between socioeconomic status (SES),academic ability, and aspirations. Other influen-tial factors include race, family size, parents'education levels, family income, academic

11 Different types of data are useful for different purposes. Cross-sectional data includes only individuals who actually enrolled in postsecondary education, and therefore canaddress the reasons students chose specific institutions but not the reasons students chose to enroll or not enroll in general. Longitudinal data, on the other hand, can beused to compare the enrollment rates of groups of students with different background characteristics at certain types of institutions, with the implication that groups withsimilar rates of enrollment have similar levels of access to a particular group of institutions.


10

aptitude and achievement, self-esteem, theaspirations of peers, high school and neighbor-hood socioeconomic status, high schoolcurriculum, and the perception of economicbenefits of college (Paulsen 1990).

� Search and application. In this phase,students decide upon a “choice set,” a group ofinstitutions that can be characterized by certainattributes such as selectivity, cost, proximity,control, level, and size. Although many institu-tions are eliminated from consideration at thisphase, it has received the least researchattention. Some differences in the timing andthe nature of college searches have been foundby race, gender, and academic aptitude. Forexample, students with higher aptitudes startthinking about college earlier and consider alarger number of schools (Paulsen 1990). Inaddition, economically disadvantaged studentsappear to be less comprehensive in theirsearches and rely more on high school counse-lors for advice (Hossler and Gallagher 1987).Institutional characteristics are importantconsiderations, especially programs, quality,cost, and location, but the priorities given to

each factor vary. Low-income students aremore likely than middle- or upper-incomestudents to rate financial assistance as impor-tant. Students with parents who have higherlevels of educational attainment are more likelyto emphasize programs and academic stan-dards and less likely to show concern aboutcosts. Higher academic ability students aremore concerned about programs and academicstandards, are aware of net price rather thanjust price, and have broader geographic limits(Paulsen 1990). (Table One.)

� Choice (selection). In this phase, individualbackground characteristics continue to play arole in influencing students' evaluation of thealternative institutions. However, the attributesof each college, offers of financial aid, andstudent preferences appear to be important aswell (Hossler and Gallagher 1987). Researchhas frequently examined the institutionalattributes that distinguish matriculants fromnon-matriculants at a particular institution,finding that the attributes that most oftendetermine where students decide to enroll arecost, financial aid, programs, location, quality,

Table One: Interaction of selected student and institutional characteristics in the search andapplication phase of the choice process

Highly selective High-cost Farther from home Private Four-yearStudents with certaincharacteristics:Gender (male) + - + - n/aRace (white) + - n/a n/a n/aParents’ Education + + + + +Family Income + + + + +Academic Aptitude + + + + +Academic Achievement + + + + n/aCollege Prep Courses (yes) + + n/a n/a n/aAspirations + + + + +

Parental Encouragement n/a n/a n/a n/a +

+ signifies a positive association • - signifies a negative association • n/a means that no association was reportedSource: Derived from summary in Paulsen 1990, pp. 55-58.

Are more likely to apply/attend institutions that are:


11

and social atmosphere (Paulsen 1990). At thesame time, student and institutional character-istics interact in determining student selection.For example, an institution becomes lessattractive to students when tuition and distancefrom home increase, especially for students atlower income levels and with lower aptitude(Paulsen 1990). Recent data from a cohort of1988 eighth graders can illustrate some ofthese interactions. Of the students whoenrolled in four-year institutions by 1994, 64percent indicated that institutional reputationwas a very important factor in their decision,44 percent cited availability of financial aid, and24 percent said level of college expenses; forthe latter, those in the lowest SES and testquartiles had higher percentages (Sanderson etal. 1996). In addition, financial aid offers wereassociated with students' decisions to attendpublic and private four-year institutions(Akerhielm et al. 1998).

These three-stage models have been broadlyaccepted, although the stages can be labeled differ-ently (Paulsen 1990). However, a three-stageprocess approach is appropriate primarily forresearch on traditional college-age students (St. John,Paulsen, and Starkey 1996). Less research attentionhas been focused on non-traditional students,despite an increase in enrollment levels (Paulsen1990). Non-traditional students, especially thosewho delay entry into postsecondary education, mayrespond to different factors and their choice pro-cesses may be less linear than those of moretraditional students.

Role of monetary factorsFinancial factors play a role at each stage of thethree-part process (St. John, Paulsen, and Starkey1996). Through parent savings and aspirations, theyaffect students’ aspirations/predisposition; theyaffect the search and application phase throughassessment of available financial resources; and theyaffect the selection of a specific college through

assessment of the net prices of alternativeinstitutions (the aspect most directly targeted bystudent aid programs).

Financial factors, however, are not the onlyvariables influencing the process, and may not evenbe the most important. One study (Manski andWise 1983) found that students with higheracademic skills (measured by class rank and SATscores), better educated parents, and higher parentincomes were more likely to apply to and attendmore expensive colleges as well as colleges withhigher average SAT scores. They also found thatcompared to academic ability and parents'education, parents’ income is a relatively unimpor-tant influence on application to college, and may bemitigated by financial aid awards.

There remains a tendency, nonetheless, for lowerSES students to enroll in different types of schoolsthan their higher SES counterparts, even afterconsidering academic factors. Non-academic factors(especially SES) still have an effect, and by them-selves explained a part of the difference in nature (interms of selectivity and per-student expenditures) ofthe postsecondary institutions that high schoolgraduates attend (Hearn 1991). “While anindividual's academic achievement is clearly a keydeterminant of college attendance, the interplay of astudent's social class background and the highschool's organizational contexts and processesappear central to the question of where an individualattends college” (McDonough 1997, 8). Studentsfeel an “entitlement” to a particular kind of collegiateeducation and are lead to organize their collegesearch around a range of “acceptable” institutions.

It also seems clear that the influence of monetaryfactors such as financial aid may differ depending onthe background characteristics of the student. Forexample, one study found that BEOG awards toupper- and middle-income youth had little effect ontheir enrollment patterns, but substantial effect on thepatterns of low-income students (Manski and Wise1983). In addition, it appears that a college becomes


12

more attractive as the availability of financial aidincreases; the effect is reduced for students at higherincome levels but enhanced for nonwhites and thoseof high academic achievement (Paulsen 1990).

Variations with definitions of choiceDespite general consensus on the factors associ-ated with stages of the choice process, theirrelative strengths may differ, depending on thetype of choice examined. For example:

� The influence of certain characteristics variesdepending on how enrollment is defined (allpostsecondary enrollment as opposed toimmediate enrollment at traditional institutions).Specifically, the influences of SES and of being ona high school track appear to decrease whenexamining the broader definition of enrollment(Hearn 1988). This is likely related to thedifferences between the choice processes oftraditional and non-traditional students.

� Changes in tuition seem to primarily affecttwo-year college attendance, where tuitionincreases discourage students from attendingcollege altogether. On the other hand,increasing proximity for two-year collegesboth encourages new college students anddiverts students from four-year colleges(Rouse 1994).

� Institutional type is associated with the prioritygiven to various factors cited by freshmen asimportant to their choice decisions. Under-graduates at four-year institutions were morelikely to cite reputation than location, price, orthe influence of others as their reason forchoosing (with students at public institutionsmore likely to choose location or price thantheir counterparts at private institutions).Beginning undergraduates at public two-yearinstitutions mentioned location as the reasonfor choosing their institution more often thanany other factors (Choy and Ottinger 1998).

13

M uch research has been conducted onfactors influencing the choice process,focusing on the perspective of students

and families. Another component of researchexamines outcomes in terms of the status ofchoice. What institutions are students with variouscharacteristics most likely to attend? Examiningwhere students are actually enrolled (outcomes)helps policymakers evaluate government policyinterventions. If differential rates of enrollment arefound, it suggests that the goal has not yet beenaccomplished. Findings also may help determinehow much more is needed to reach the goal ofcollege choice, however defined.

Each specific conceptualization of choice has implica-tions for measuring outcomes. For example, if accesswere defined as access to any postsecondary institu-tion, then the basic “institution of access” would bethe public two-year institution, and choice wouldreflect a comparison of enrollment patterns at publictwo-year institutions to enrollment at other institu-tions, especially four-year institutions. Indicators ofchoice under other definitions might compareenrollment at public versus private institutions,higher-priced versus lower-priced institutions, orselective versus non-selective institutions.12 Inexploring the role of financial aid (or, specifically,federal need-based aid), outcomes measures oftenassume that, all else being equal, a student mightprefer to attend a more expensive institution over aless expensive institution if he/she could afford to.

This assumption accounts for policy efforts thatenable a qualified student to attend a moreexpensive institution if desired.

There are some limitations to outcomes-basedanalyses, many of which look only at studentsenrolled in postsecondary institutions at a certainpoint in time. In addition, the analyses tend toframe differences in terms of enrollment at institu-tions with varying price levels, which implies thatmonetary factors are the only forces at work. Asmentioned previously, however, individual studentpreferences and other academic and institutionalfactors affect enrollment decisions throughout thechoice process. Analysis of outcomes only indirectlycaptures the effect of the factors that narrow therange of institutions considered before theselection decision is made. In other words, mostoutcomes measures reflect the cumulativeinfluences of a range of factors, and it is difficult toseparate out distinct influences.13 Yet aggregateoutcomes are one of the tools used to judge theeffectiveness of federal aid programs in meetingtheir goals. It is best to keep this apparent inconsis-tency in mind while examining the results.

Another important qualification relates specificallyto studies of enrollment at two-year versus four-year institutions. Clearly, some students whobegin at two-year institutions later transfer tofour-year institutions, yet these movements arenot captured in cross-sectional studies of

outcomesthe status of choice in a policy context

12 The ability to “choose” more selective institutions is tied closely to academic qualifications as well as financial ability, perhaps more so than the other definitions.13 In particular, there is at least a perceived correlation between institutional quality (selectivity) and price, such that differences in enrollment between higher-priced and

lower-priced institutions might also be reflecting academic factors.


14

enrollment differences. Broadly, outcomes mea-sures reflect the choices of students at a particularpoint in time, not their future goals or theirprevious postsecondary experiences.

Despite these caveats, researchers have attemptedto investigate the status of choice and the effective-ness of policy through a review of enrollment atdifferent types of institutions by students fromvarying economic backgrounds. The followingsections are based on previous research models,but use original analysis of recent data to updateand expand on these models.14 In some cases, thenew analyses are supplemented by previouslongitudinal research. Each section examines thestatus of choice from a slightly different perspec-tive; taken together, these perspectives can help usbetter understand college choice.

Distribution of students accordingto institutional typeAt the most basic level, status of choice can beexamined by looking at the enrollment of studentswith different backgrounds by institutional type.This analysis assumes that if choice is being fullyachieved, the distribution of students with differentcharacteristics will be somewhat equal. Put anotherway, if low family income, first-generation status,and delayed enrollment characteristics are notbarriers to student decisions at some point in thechoice process, then students exhibiting thesecharacteristics should be just as likely as their moreadvantaged counterparts to attend various types ofinstitutions. The first level of analysis comprises thistype of outcomes data. Differential rates of enroll-ment do not necessarily mean that certain studentsdo not have choice, but rather show that certaingroups of students are more likely to attend certaintypes of institutions. The results may provide

ground for further exploration—if the goal is equalrates of enrollment for those groups of students.

Overall distribution of first-year undergraduates byinstitutional type can set the context for differencesin enrollment. In 1999-2000, 62 percent of all first-year undergraduates were enrolled at publictwo-year institutions; 19 percent attended publicfour-year institutions, 9 percent attended privatenot-for-profit four-year institutions, 8 percentattended proprietary schools, and the rest (2percent) attended other types of institutions(NCES, 1999-2000).

Not surprisingly, previous studies have found thatdistribution of students among types of institutionsdiffered according to student characteristics:certain types of students appear to be more or lesslikely to attend certain types of institutions. Oneexample is an analysis of 1995-96 data by Choy andOttinger (1998), which examined student enroll-ment from two complementary perspectives: thepercentage of beginning postsecondary studentswith specific characteristics who were enrolled atdifferent types of institutions, and the percentage ofstudents with various characteristics enrolled atspecific types of institutions. The analysis comparedcolleges by level (four-year, two-year, and less-than-two-year) and by control (public, privatenot-for-profit, and private for-profit). Tables Twoand Three replicate the analysis for first-yearundergraduates using more recent data, withsimilar conclusions.15

In 1999-2000, where freshmen were enrolled wasrelated to whether they had “traditional” or “non-traditional” characteristics. Freshmen can becharacterized as non-traditional if they have at leastone of the following characteristics: enrolling parttime, being financially independent, being a single

14 Data are drawn from the National Postsecondary Student Aid Study (NPSAS), for the years 1999-2000 and 1989-1990.15 Choy and Ottinger (1998) looked at beginning postsecondary students, while the following analysis examines first-year undergraduates (freshmen), which may include some

students who began their postsecondary education in previous years but are still considered first-year students.


15

Tabl

e Tw

o: P

erce

ntag

e di

strib

utio

n of

firs

t-yea

r und

ergr

adua

tes

acco

rdin

g to

inst

itutio

nal t

ype,

199

9-20

00

Pri

vate

not

-Pri

vate

not

-Pub

lic les

sPub

licPub

licfo

r-pr

ofit

les

sfo

r-pr

ofit

Pri

vate

Tota

lth

an 2

-yea

r 2

-yea

r4-y

ear

than

4-y

ear

4-y

ear

for

prof

it

Tota

l1

00

.0%

0.8

%6

2.0

%1

8.9

%1

.2%

9.2

%8

.0%

STU

DEN

T C

HA

RA

CTE

RIS

TIC

SG

ende

r

Mal

e1

00

.0%

0.7

%6

1.7

%2

0.0

%1

.2%

9.3

%7

.2%

F

emal

e1

00

.0%

0.8

%6

2.2

%1

8.0

%1

.2%

9.2

%8

.7%

Age

as

of 1

2/3

1/9

9

18

yea

rs o

r yo

unge

r1

00

.0%

0.2

%4

1.4

%3

5.8

%1

.2%

18

.0%

3.4

%

19

-23

yea

rs1

00

.0%

0.6

%6

1.0

%2

0.3

%1

.6%

8.3

%8

.1%

2

4 y

ears

or

olde

r1

00

.0%

1.2

%7

5.1

%7

.5%

0.9

%4

.8%

10

.6%

Rac

e-et

hnic

ity

W

hite

, no

n-H

ispa

nic

10

0.0

%0

.6%

62

.3%

20

.4%

1.1

%9

.5%

6.2

%

Bla

ck,

non-

His

pani

c1

00

.0%

1.6

%6

1.7

%1

5.5

%1

.6%

8.2

%1

1.5

%

His

pani

c or

Lat

ino

10

0.0

%0

.6%

60

.0%

15

.5%

1.7

%1

0.0

%1

2.4

%

Asi

an/

Pac

ific

Isla

nder

10

0.0

%0

.7%

61

.0%

21

.5%

0.8

%6

.9%

9.1

%

Am

eric

an In

dian

/A

lask

a N

ative

10

0.0

%0

.4%

74

.6%

10

.6%

1.2

%7

.1%

6.1

%

Sin

gle

pare

nt s

tatu

s

Sin

gle

pare

nt1

00

.0%

1.4

%6

7.3

%8

.3%

1.6

%5

.1%

16

.4%

Dep

ende

ncy

stat

us f

or f

inan

cial

aid

D

epen

dent

10

0.0

%0

.4%

51

.3%

29

.0%

1.4

%1

3.5

%4

.4%

In

depe

nden

t w

ithou

t de

pend

ent

10

0.0

%0

.9%

74

.7%

7.9

%0

.8%

4.5

%1

1.2

%

Inde

pend

ent

with

dep

ende

nt1

00

.0%

1.3

%7

2.6

%8

.0%

1.1

%4

.7%

12

.4%

Tota

l in

com

e by

dep

ende

ncy

stat

us

Dep

ende

nt,

less

tha

n $

30

,00

01

00

.0%

0.3

%5

5.1

%2

4.8

%1

.5%

10

.7%

7.6

%

Dep

ende

nt,

$3

0,0

00

to

$6

9,9

99

10

0.0

%0

.6%

53

.8%

28

.8%

1.4

%1

1.6

%3

.9%

D

epen

dent

, $

70

,00

0 o

r m

ore

10

0.0

%0

.2%

45

.1%

32

.7%

1.4

%1

8.2

%2

.4%

Tota

l in

com

e by

dep

ende

ncy

stat

us

Inde

pend

ent,

less

tha

n $

20

,00

01

00

.0%

1.3

%6

4.9

%8

.7%

1.6

%4

.9%

18

.6%

In

depe

nden

t, $

20

,00

0 t

o $

49

,99

91

00

.0%

1.1

%7

7.8

%7

.2%

0.7

%3

.6%

9.6

%

Inde

pend

ent:

$5

0,0

00

or

mor

e1

00

.0%

0.9

%8

0.8

%7

.9%

0.5

%5

.6%

4.4

%


16

Tabl

e Tw

o: P

erce

ntag

e di

strib

utio

n of

firs

t-yea

r und

ergr

adua

tes

acco

rdin

g to

inst

itutio

nal t

ype,

199

9-20

00

Par

ents

’ hig

hest

edu

cation

lev

el

HS o

r le

ss1

00

.0%

1.0

%6

7.3

%1

4.1

%1

.3%

6.3

%1

0.1

%

Som

e PSE

10

0.0

%0

.7%

63

.7%

19

.3%

1.1

%7

.7%

7.6

%

Bac

helo

r’s

degr

ee1

00

.0%

0.7

%5

8.6

%2

3.6

%1

.1%

10

.5%

5.6

%

Adv

ance

d de

gree

10

0.0

%0

.3%

44

.6%

30

.1%

1.1

%1

8.9

%5

.0%

Res

pond

ent

has

depe

nden

ts

No

depe

nden

ts1

00

.0%

0.5

%5

7.3

%2

3.6

%1

.3%

11

.2%

6.1

%

Has

dep

ende

nts

10

0.0

%1

.3%

72

.6%

8.0

%1

.1%

4.7

%1

2.4

%

Del

ayed

enr

ollm

ent

into

PSE

N

o de

lay

10

0.0

%0

.3%

53

.5%

26

.4%

1.2

%1

2.7

%5

.8%

D

elay

ed e

nrol

lmen

t1

00

.0%

1.2

%7

2.1

%1

0.0

%1

.2%

4.8

%1

0.7

%

Hig

h sc

hool

deg

ree

H

igh

scho

ol d

iplo

ma

10

0.0

%0

.7%

60

.6%

20

.3%

1.2

%9

.7%

7.5

%

GED

/ot

her/

none

10

0.0

%1

.3%

73

.9%

6.2

%1

.6%

4.6

%1

2.5

%

EN

RO

LLM

EN

T, F

INA

NC

IAL

AID

,A

ND

WO

RK

CH

AR

AC

TER

ISTI

CS

Deg

ree

Pro

gram

C

ertif

icat

e1

00

.0%

3.4

%6

6.2

%4

.0%

1.7

%2

.0%

22

.7%

A

ssoc

iate

’s d

egre

e1

00

.0%

0.0

%8

6.0

%4

.9%

1.7

%2

.8%

4.6

%

Bac

helo

r’s

degr

ee1

00

.0%

0.0

%2

.1%

64

.3%

0.0

%3

0.9

%2

.7%

N

o un

derg

radu

ate

degr

ee1

00

.0%

0.0

%8

3.3

%1

2.4

%0

.1%

3.5

%0

.7%

Att

enda

nce

inte

nsity

E

x clu

siv e

ly ful

l-tim

e1

00

.0%

1.0

%4

2.2

%2

7.7

%1

.9%

14

.3%

12

.9%

P

art

time

or m

ixed

10

0.0

%0

.5%

81

.9%

10

.0%

0.5

%4

.1%

3.1

%

Pri

vate

not

-P

riva

te n

ot-

Public

les

sP

ublic

Public

for-

pro

fit

less

for-

pro

fit

Pri

vate

Tota

lth

an 2

-yea

r 2

-yea

r4

-yea

rth

an 4

-yea

r4

-yea

rfo

r pr

ofit


17

Aid

tot

al a

mou

nt

No

aid

10

0.0

%0

.8%

79

.1%

13

.5%

0.6

%4

.1%

1.9

%

Aid

ed1

00

.0%

0.7

%4

8.5

%2

3.0

%1

.7%

13

.2%

12

.8%

Pri

mar

y ro

le-s

tude

nt o

r em

ploy

ee

Stu

dent

who

wor

ks1

00

.0%

0.6

%5

7.1

%2

2.2

%1

.4%

11

.1%

7.5

%

Em

ploy

ee w

ho s

tudi

es1

00

.0%

0.9

%7

7.8

%8

.8%

0.7

%4

.8%

7.1

%

Doe

s no

t w

ork

10

0.0

%0

.8%

46

.1%

28

.5%

1.6

%1

3.1

%1

0.0

%

Wor

k in

tens

ity

whi

le e

nrol

led

D

id n

ot w

ork

10

0.0

%0

.8%

46

.1%

28

.5%

1.6

%1

3.1

%1

0.0

%

Wor

ked

part

-tim

e1

00

.0%

0.7

%5

5.9

%2

3.0

%1

.4%

11

.7%

7.4

%

Wor

ked

full-

time

10

0.0

%0

.7%

76

.5%

9.8

%0

.8%

5.0

%7

.2%

Note

: Det

ails

may

not

add

to to

tals

due

to ro

undi

ng. •

Sou

rce:

NCES

199

9-20

00, b

ased

on

Choy

and

Ottin

ger 1

997

Tabl

e Tw

o: P

erce

ntag

e di

strib

utio

n of

firs

t-yea

r und

ergr

adua

tes

acco

rdin

g to

inst

itutio

nal t

ype,

199

9-20

00

Pri

vate

not

-Pri

vate

not

-Pub

lic les

sPub

licPub

licfo

r-pr

ofit

les

sfo

r-pr

ofit

Pri

vate

Tota

lth

an 2

-yea

r 2

-yea

r4-y

ear

than

4-y

ear

4-y

ear

for

prof

it


18

Tabl

e Th

ree:

Per

cent

age

dist

ribut

ion

of fi

rst-y

ear u

nder

grad

uate

s ac

cord

ing

to s

tude

nt c

hara

cter

istic

s, 1

999-

2000

Priva

te n

ot-

Priva

te n

ot-

Priva

teTo

tal

Pub

lic le

ssPub

licPub

licfo

r-pr

ofit le

ssfo

r-pr

ofit

for-

than

2-y

ear

2-y

ear

4-y

ear

than

4-y

ear

4-y

ear

prof

it

Tota

l1

00

.0%

10

0.0

%1

00

.0%

10

0.0

%1

00

.0%

10

0.0

%1

00

.0%

STU

DEN

T C

HA

RA

CTE

RIS

TIC

SG

ende

r

Mal

e4

4.3

%4

3.6

%4

4.1

%4

6.9

%4

4.1

%4

4.5

%3

9.5

%

Fem

ale

55

.8%

56

.5%

56

.0%

53

.1%

55

.9%

55

.5%

60

.5%

Age

as

of 1

2/3

1/9

9

18

yea

rs o

r yo

unge

r2

3.7

%6

.0%

15

.8%

44

.9%

23

.2%

46

.3%

10

.1%

1

9-2

3 y

ears

36

.5%

29

.4%

36

.0%

39

.4%

48

.5%

33

.0%

37

.1%

2

4 y

ears

or

olde

r3

9.8

%6

4.7

%4

8.2

%1

5.7

%2

8.3

%2

0.7

%5

2.8

%

Rac

e-et

hnic

ity

W

hite

, no

n-H

ispa

nic

63

.9%

50

.4%

64

.2%

69

.0%

56

.1%

65

.6%

49

.6%

B

lack

, no

n-H

ispa

nic

14

.1%

29

.9%

14

.1%

11

.6%

18

.4%

12

.5%

20

.2%

H

ispa

nic

or L

atin

o1

3.9

%1

1.0

%1

3.5

%1

1.4

%1

8.8

%1

5.0

%2

1.5

%

Asi

an/

Pac

ific

Isla

nder

5.3

%5

.2%

5.2

%6

.0%

3.6

%4

.0%

6.0

%

Am

eric

an In

dian

/A

lask

a N

ative

1.0

%0

.5%

1.1

%0

.5%

0.9

%0

.7%

0.7

%

Sin

gle

pare

nt s

tatu

s

Sin

gle

pare

nt1

4.3

%2

7.6

%1

5.5

%6

.3%

18

.1%

7.9

%2

9.2

%

Not

sin

gle

pare

nt8

5.7

%7

2.4

%8

4.5

%9

3.7

%8

1.9

%9

2.1

%7

0.8

%

Dep

ende

ncy

stat

us f

or f

inan

cial

aid

D

epen

dent

51

.9%

27

.3%

43

.0%

79

.8%

60

.5%

76

.0%

28

.4%

In

depe

nden

t w

ithou

t de

pend

ent

17

.9%

21

.5%

21

.6%

7.5

%1

1.4

%8

.7%

25

.1%

In

depe

nden

t w

ith d

epen

dent

30

.2%

51

.2%

35

.4%

12

.8%

28

.1%

15

.3%

46

.5%

Tota

l in

com

e by

dep

ende

ncy

stat

us

Dep

ende

nt,

less

tha

n $

30

,00

01

3.5

%6

.0%

12

.0%

17

.7%

17

.0%

15

.7%

12

.7%

D

epen

dent

, $

30

,00

0 t

o $

69

,99

92

1.7

%1

6.5

%1

8.9

%3

3.1

%2

4.7

%2

7.3

%1

0.5

%

Dep

ende

nt,

$7

0,0

00

or

mor

e1

6.7

%4

.9%

12

.2%

29

.0%

18

.8%

33

.1%

5.1

%

Inde

pend

ent,

less

tha

n $

20

,00

01

8.9

%3

2.2

%1

9.8

%8

.7%

25

.1%

10

.1%

43

.8%

In

depe

nden

t, $

20

,00

0 t

o $

49

,99

91

8.1

%2

7.8

%2

2.8

%6

.9%

9.8

%7

.1%

21

.7%

In

depe

nden

t: $

50

,00

0 o

r m

ore

11

.1%

12

.7%

14

.4%

4.6

%4

.6%

6.7

%6

.1%

Par

ents

’ hig

hest

edu

cation

lev

el

HS o

r le

ss4

2.1

%5

5.1

%4

6.1

%3

0.5

%4

5.1

%2

8.6

%5

4.1

%

Som

e PSE

23

.4%

21

.5%

24

.2%

23

.1%

21

.7%

19

.5%

22

.6%

B

ache

lor’s

degr

ee2

0.8

%1

7.7

%1

9.8

%2

5.2

%2

0.0

%2

3.7

%1

4.7

%

Adv

ance

d de

gree

13

.7%

5.7

%9

.9%

21

.2%

13

.2%

28

.1%

8.7

%


19

Tabl

e Th

ree:

Per

cent

age

dist

ribut

ion

of fi

rst-y

ear u

nder

grad

uate

s ac

cord

ing

to s

tude

nt c

hara

cter

istic

s, 1

999-

2000

Priva

te n

ot-

Priva

te n

ot-

Priva

tePub

lic le

ssPub

licPub

licfo

r-pr

ofit le

ssfo

r-pr

ofit

for-

Tota

lth

an 2

-yea

r2-y

ear

4-y

ear

than

4-y

ear

4-y

ear

prof

it

Res

pond

ent

has

depe

nden

ts

No

depe

nden

ts6

9.8

%4

8.8

%6

4.6

%8

7.2

%7

1.9

%8

4.7

%5

3.5

%

Has

dep

ende

nts

30

.2%

51

.2%

35

.4%

12

.8%

28

.1%

15

.3%

46

.5%

Del

ayed

enr

ollm

ent

into

PSE

N

o de

lay

55

.4%

25

.6%

48

.0%

76

.8%

56

.5%

76

.6%

40

.0%

D

elay

ed e

nrol

lmen

t4

4.6

%7

4.4

%5

2.0

%2

3.2

%4

3.6

%2

3.4

%6

0.0

%

Hig

h sc

hool

deg

ree

H

igh

scho

ol d

iplo

ma

90

.1%

83

.1%

88

.2%

96

.8%

87

.5%

95

.0%

84

.7%

G

ED/

othe

r/no

ne9

.9%

16

.9%

11

.8%

3.2

%1

2.6

%5

.0%

15

.3%

EN

RO

LLM

EN

T , F

INA

NCIA

L A

ID,

AN

D W

OR

K C

HA

RA

CTE

RIS

TIC

SD

egre

e pr

ogra

m

Cer

tific

ate

22

.0%

99

.5%

23

.5%

4.7

%3

0.1

%4

.9%

62

.4%

A

ssoc

iate

’s d

egre

e5

0.9

%0

.6%

70

.6%

13

.3%

69

.7%

15

.4%

29

.4%

B

ache

lor’s

degr

ee2

3.3

%0

.0%

0.8

%7

9.5

%0

.1%

78

.3%

7.9

%

No

unde

rgra

duat

e de

gree

3.8

%0

.0%

5.1

%2

.5%

0.1

%1

.5%

0.3

%

Att

enda

nce

inte

nsity

Ex

clus

ivel

y fu

ll-tim

e5

0.3

%6

4.6

%3

4.3

%7

3.7

%7

9.0

%7

8.1

%8

1.1

%

Par

t tim

e or

mixed

49

.8%

35

.4%

65

.7%

26

.3%

21

.0%

21

.9%

18

.9%

Aid

tot

al a

mou

nt

No

aid

43

.9%

46

.8%

56

.1%

31

.5%

21

.6%

19

.6%

10

.2%

A

ided

56

.1%

53

.2%

43

.9%

68

.5%

78

.4%

80

.4%

89

.8%

Pri

mar

y ro

le-s

tude

nt o

r em

ploy

ee

Stu

dent

who

wor

ks4

0.2

%3

1.2

%3

7.1

%4

7.4

%4

7.5

%4

8.1

%3

8.2

%

Empl

oyee

who

stu

dies

36

.2%

43

.7%

45

.4%

16

.9%

21

.9%

18

.6%

32

.2%

D

oes

not

wor

k2

3.6

%2

5.1

%1

7.6

%3

5.8

%3

0.6

%3

3.3

%2

9.7

%

Wor

k in

tens

ity

whi

le e

nrol

led

D

id n

ot w

ork

23

.5%

24

.9%

17

.4%

35

.5%

30

.4%

33

.2%

29

.6%

W

orke

d pa

rt-tim

e3

5.3

%3

4.8

%3

1.8

%4

3.1

%4

0.9

%4

4.5

%3

2.9

%

Wor

ked

full-

time

41

.2%

40

.3%

50

.8%

21

.4%

28

.7%

22

.2%

37

.5%

Note

: Det

ails

may

not

add

to to

tals

due

to ro

undi

ng.

• S

ourc

e: NC

ES 1

999-

2000

, bas

ed o

n Ch

oy an

d Ot

tinge

r 199

7


20

parent, being older, delaying enrollment, having aGED or no high school diploma, having dependentsother than a spouse, and working full time whileenrolled (Horn 1996). In general, students withnon-traditional characteristics were more likely toattend two-year institutions and proprietaryinstitutions16 than were traditional freshmen. (TableTwo.) For example:

� Older students, independent students, studentswith dependents, students who delayed enroll-ment, first-generation students, and studentswithout a high school diploma were more likelyto attend public two-year or proprietary schoolsthan their counterparts. For example, 75percent of students 24 years or older wereenrolled at public two-year institutions, com-pared to 41 percent of students 18 years oryounger. Students who worked full-time whileenrolled and students who attended on a parttime or part year basis were more likely toattend public 2-year institutions.17

Other differences also are clear:

� The lowest-income dependent students weremore likely to attend public two-year institu-tions (55 percent) and proprietaryinstitutions (8 percent) than the highest-income dependent students (45 and 2percent, respectively). A different patternwas found for independent students, wherethe lowest-income students were more likelyto attend proprietary schools, but the high-est-income students were more likely toattend public two-year institutions.

� Students who did not receive financial aid in1999-2000 were more likely than their coun-terparts who received aid to be attendingpublic two-year institutions, 79 percent com-pared to 49 percent.

� Black and Hispanic students were more likelyto attend proprietary schools than werestudents of other races/ethnicities, 12 percentfor Black and Hispanic students compared to 6percent for white students.

Conversely, in 1999-2000, different types ofinstitutions tended to serve students with differentcharacteristics. (Table Three) In general, this brokedown along the lines of four-year versus less-than-four-year schools:

� Public two-year institutions and proprietaryschools tend to serve older students, financiallyindependent students, first generation students,and higher proportions of single parents thanfour-year schools. However, there were somedifferences in the types of students served.Proprietary schools enrolled a higher proportionof Black and Hispanic students, as well as studentsattending on a full-time basis, than public two-year institutions. In addition, public two-yearinstitutions enroll students from a range ofincome backgrounds, while proprietary schoolsprimarily serve students from a low-incomebackground. Public two-year institutions primarilyserve students enrolled in an associate's degreeprogram, while proprietary schools and publicless-than-two-year institutions primarily enrollstudents in certificate programs.

� Public and private not-for-profit four-yearinstitutions tend to serve more traditionalfirst-year students, i.e., younger students,dependent students with middle or upperincome backgrounds, students in bachelor' sdegree programs, and students attending on afull-time basis. At both types of institutions,the majority of students received financial aid,although the proportion was slightly larger atprivate not-for-profit four-year institutions.

16 Although some proprietary schools are four-year institutions, for the purposes of this analysis they are classified with public two-year institutions because the overwhelm-ing majority of their programs are for certificates or associate’s degrees.

17 Most of the differences found regarding students with non-traditional characteristics, etc. also hold when examining the distribution of students according to the tuitionlevel of the institution they attend. In other words, non-traditional students are more likely to attend less expensive schools.


21

Lowest Lower Upper Highest1999-2000 quintile middle Middle middle quintile

Private not-for-profit PhD 2.4% 2.9% 2.6% 3.3% 6.8%Private not-for-profit non-PhD 5.8% 4.5% 5.6% 5.7% 6.5%Private not-for-profit 2-year 1.7% 1.3% 1.0% 1.1% 0.9%All private not-for-profit 9.9% 8.8% 9.3% 10.1% 14.2%

Private for profit 14.7% 9.9% 7.0% 5.2% 3.2%All private 24.6% 18.6% 16.3% 15.3% 17.4%

Public 4-year PhD 8.7% 9.7% 10.0% 11.3% 13.6%Public 4-year non-PHD 9.6% 7.8% 7.8% 7.8% 7.4%Public 2-year or less 57.1% 63.9% 66.0% 65.6% 61.6%All public 75.4% 81.4% 83.7% 84.7% 82.6%

All 4-year 26.5% 24.9% 26.0% 28.1% 34.4%All 2-year/proprietary 73.5% 75.1% 74.0% 71.9% 65.7%

Total 100.0% 100.0% 100.0% 100.0% 100.0%




Public 4-year PhD 8.0% 8.4% 9.5% 9.8% 11.1%Public 4-year non-PhD 8.7% 8.3% 8.6% 9.1% 7.8%Public 2-year or less 44.6% 54.3% 58.3% 60.9% 61.5%All public 61.3% 70.9% 76.4% 79.8% 80.5%


Total 100.0% 100.0% 100.0% 100.0% 100.0%

Lowest Lower Upper HighestPercentage point change: quintile middle Middle Middle quintile

Private not-for-profit PhD -0.3 0.7 0.3 0.2 2.0Private not-for-profit non-PhD 0.1 -0.2 0.3 -0.2 -1.0Private not-for-profit 2-year -1.1 -0.8 -1.5 -0.9 -0.5All private not-for-profit -1.3 -0.3 -0.8 -0.9 0.5

Private for profit -12.8 -10.2 -6.5 -3.9 -2.7All private -14.1 -10.4 -7.3 -4.8 -2.2

Public 4-year PhD 0.7 1.2 0.5 1.5 2.5Public 4-year non-PHD 0.9 -0.5 -0.8 -1.3 -0.4Public 2-year or less 12.5 9.7 7.6 4.7 0.1All public 14.1 10.5 7.3 4.8 2.2

All 4-year 1.4 1.3 0.3 0.2 3.1All 2-year/proprietary -1.4 -1.3 -0.3 -0.2 -3.1

Note: Excludes those who attended multiple institutions. Details may not add to totals due to rounding.Source: NCES 1999-2000 and 1989-90, based on McPherson and Schapiro 1998 and 1999

Table Four: Distribution of freshman enrollment by income quintile acrossinstitutional types, 1989-90 and 1999-2000


22

Table Five: Distribution of full-time, full-year, dependent freshman enrollment by incomequintile across institutional types, 1989-90 and 1999-2000






Total 100.0% 100.0% 100.0% 100.0% 100.0%

Lowest Lowest Upper Highest1989-90 quintile middle Middle middle quintile





Total 100.0% 100.0% 100.0% 100.0% 100.0%

Lowest Lowest Upper HighestPercentage point change: quintile middle Middle middle quintile

Private not-for-profit PhD -0.6 2.2 1.7 1.7 2.7Private not-for-profit non-PhD -2.9 -2.2 0.5 -1.3 -3.9Private not-for-profit 2-year -0.4 -0.3 -0.2 0.2 -0.6All private not-for-profit -3.8 -0.2 2.0 0.6 -1.8


Public 4-year PhD -2.6 4.6 1.2 4.1 3.8Public 4-year non-PHD -4.8 -3.2 -6.2 -5.3 -8.8Public 2-year or less 12.6 3.0 5.5 1.8 7.3All public 5.2 4.4 0.5 0.6 2.2

All 4-year -10.9 1.5 -2.8 -0.8 -6.3All 2-year/proprietary 10.8 -1.4 2.8 0.8 6.3

Note: Excludes those who attended multiple institutions. Details may not add to totals due to rounding. • Source: NCES 1999-2000 and 1989-90, based onMcPherson and Schapiro 1998 and 1999


23

Private not-for-profit four-year institutionswere slightly more likely to enroll studentswhose parents had an advanced degree.

Longitudinal research studies have found similardifferences in the characteristics of studentsattending various types of institutions.18 Ultimately,it appears that the greatest difference in thedistribution of undergraduates may occur at theextremes—the lowest-income students are con-centrated in the lowest-priced institutions, and thehighest-income students dominate at the highest-priced institutions (Lee 1999).

Increasing stratification over timeThe next level of analysis regarding the status ofchoice looks at the issue of change over time, andwhether public policies appear to be helping toencourage an increasing level of choice for all stu-dents. What kinds of shifts have occurred over thelast decade, in terms of the proportions of studentsfrom certain income backgrounds that are attendingspecific types of institutions? The higher educationcommunity is concerned about the concentration ofdisadvantaged students in certain institutions movingin the wrong direction—that is, rather than seeing animprovement in choice as a result of public policy,choice actually may be declining for disadvantagedstudents as tuitions increase and financial aid (espe-cially federal need-based grants) fails to keep pace.Thus, higher education may be becoming increas-ingly stratified by income.

One recent study raised this issue after ananalysis of changes in freshmen enrollmentpatterns between 1981 and 1998 (McPhersonand Schapiro 1998 and 1999).19 The study foundthat lower-income students were clustered incommunity colleges and that the likelihood of astudent attending a four-year institution (espe-cially a private one) depends on their parents’income. Between 1981 and 1998, theproportion of lower- and lower-middle-incomestudents attending public two-year institutionsincreased, while the proportions of all otherincome groups fell. At the same time, theproportion of upper-income students attendinguniversities increased over time, the proportionattending less selective, private four-yearcolleges decreased, and the proportion attend-ing public four-year institutions increased.

To replicate the analysis, changes over the decadebetween 1989-90 and 1999-2000 were examined.20

Broad changes in student characteristics andfinancing occurred over this period—for example,the definition of independent student changed in1993, which made it more difficult to qualify as anindependent student (see box on page 25). Although the cross-sections of the two years arenot strictly comparable, it is useful to look at thedistribution of freshmen enrollment at the begin-ning of the decade and at its end.

In 1999-2000, the majority of freshmen in all incomequintiles attended public two-year institutions. (Table

18 For example, a longitudinal study of 1988 eighth graders found that of those who enrolled by 1994, almost 50 percent of Hispanics enrolled in public two-year institutions,and that higher percentages of students in the higher SES and test quartiles enrolled in private four-year institutions (Sanderson et al. 1996). Among 1988 eighth graderswho enrolled by 1994, low-income students were less likely to attend four-year institutions than higher-income students, even among high test score students, and low-income students were disproportionately found in public institutions, even within the group of high test score students (Akerhielm et al. 1998).

19 McPherson and Schapiro (1998 and 1999) used self-reported survey data of first-time, full-time freshman, from a dataset in which relatively few communitycolleges are represented.

20 The benefit of the dataset used by McPherson and Schapiro (1998 and 1999) is that they were able to group institutions by selectivity. However, the NPSAS data for 1999-2000 does not allow this as well. In the following analysis, “PhD institutions” were used as a proxy for more selective institutions (as opposed to “non-PhD institutions”).Although this is an imperfect proxy, the data reveal that PhD institutions tend to be more expensive than non-PhD institutions, and their students tend to have higher testscores. In 1999-2000, the average tuition and fees faced by first-year dependent undergraduates attending public PhD-granting institutions were $3,980, compared with$2,729 at public non-PhD institutions (the respective figures were $17,035 and $11,428 at private not-for-profit institutions). In addition, the average SAT combined scorepercentile rank for first-year dependent undergraduates attending public PhD-granting institutions was 60, compared to 41 at public non-PhD institutions (the respectivefigures were 69 and 60 at private not-for-profit institutions).


24






Total 100.0% 100.0% 100.0% 100.0% 100.0%

Lowest Lowest Upper Highest1989-90 quintile middle Middle middle quintile





Total 100.0% 100.0% 100.0% 100.0% 100.0%

Lowest Lowest Upper HighestPercentage point change: quintile middle Middle middle quintile

Private not-for-profit PhD -1.2 0.2 -0.3 -0.7 0.1Private not-for-profit non-PhD 1.5 1.0 -0.5 0.4 -0.7Private not-for-profit 2-year -0.4 -0.2 -1.1 -1.4 -0.5All private not-for-profit -0.1 0.9 -2.0 -1.7 -1.2


Public 4-year PhD -0.6 -0.9 -0.7 0.5 -0.3Public 4-year non-PHD 0.8 -0.3 1.1 0.4 1.0Public 2-year or less 6.5 3.7 3.8 1.5 1.9All public 6.6 2.5 4.2 2.4 2.6

All 4-year 0.5 0.0 -0.5 0.6 0.1All 2-year/proprietary -0.5 0.0 0.4 -0.6 -0.1

Note: Excludes those who attended multiple institutions. Details may not add to totals due to rounding. • Source: NCES 1999-2000 and 1989-90, basedon McPherson and Schapiro 1998 and 1999

Table Six: Distribution of part time, part-year, or mixed independent freshman enrollment byincome quintile across institutional types, 1989-90 and 1999-2000


25

In examining the changes in students’ enrollmentdistribution over the decade, it is important to keep in

mind several changes in public policies, competitivepressures, and the types of students enrolling inpostsecondary education over this period.

Transition to college. More students enrolled inhigher education over this period. Continuing the trendfrom previous decades, college enrollment rates of highschool graduates increased slightly over the decade,from 60 percent in 1990 to 63 percent in 1999. Inparticular, the enrollment rates of African-American highschool graduates increased, from 46 percent to 59percent (NCES 2002).

Overall, total fall enrollment in higher educationinstitutions increased by about 4 percent between 1990and 1997. However, the number of first-time freshmendeclined between 1990 and 1997, by about 5 percent(NCES 2002).

Composition of students. The “traditional” stu-dent—dependent, attending full-time, and working nomore than part-time—is no longer typical ofpostsecondary education. By 1999-2000, three-quartersof all undergraduates had at least one “non-traditional”characteristic (delayed enrollment, part-time atten-dance, working full-time, financially independent,single parent, has dependents, or no high schooldiploma), although the proportion varies by type ofinstitution (NCES 2002b).

Although the changes were not as drastic as in previousdecades, the types of students enrolling in postsecondaryeducation have changed over the past decade. Thepercentage of first-year undergraduates who were non-white increased from 28 percent in 1989-90 to 36 percentin 1999-2000, for example, and the percentage who weresingle parents increased from 10 percent to 14 percent(NCES 1989-90 and 1999-2000).

At the same time, the percentage of first-yearundergraduates who were classified as dependentstudents increased, from 46 percent to 52 percent(NCES 1989-90 and 1999-2000). The definition ofan independent student (for federal financial aidpurposes) changed through the 1992 HigherEducation Act (HEA) amendments, however, so itwas expected that fewer students would qualify asfinancially independent.

Financial aid policy. Over the period 1989-90 to1999-2000, a higher proportion of students tended toreceive financial aid, whether from government,institutional, or private sources. In fact, the percent-age of first-year undergraduates who received aidfrom any source increased from 41 percent to 56percent over this period (NCES 1989-90 and 1999-2000). At the same time, there were shifts in the typeof aid awarded to postsecondary students. Forexample, in 1989-90 grants made up 48 percent oftotal aid to students and loans made up 49 percent;by 1999-2000, the percentage for grants had de-creased to 40 percent of all aid, while loans hadincreased to 59 percent (College Board 2000).

Meanwhile, the 1992 reauthorization of the HEAaffected the need analysis methodology used toallocate federal, need-based student aid. For dependentstudents in particular, these changes included: 1)exclusion of home equity as an asset; 2) some low-income parents were no longer required to showassets; 3) the minimum contribution from students waseliminated; and 4) the amount expected from studentearnings was reduced. All of these changes workedtoward lower EFCs and a larger percentage of studentsqualifying for need-based aid. The changes relevant toindependent students were more mixed.

The reauthorization and subsequent legislation duringthis period also enacted increases in the maximum

The context of postsecondary education in the 1990s


26

Four.) In general, however, students from higherincome quintiles were more likely to attend four-yearinstitutions (particularly PhD-granting institutions),while students from lower income quintiles weremore likely to attend proprietary schools. Overtime—since 1989-90—the primary shift appears tohave been away from proprietary schools and towardpublic two-year institutions, especially among thelower income quintiles. These trends change, how-ever, for different types of freshmen, based ondependency and attendance status.

Slightly more than half (52 percent) of all first-yearundergraduates were dependent students in1999-2000; of these students, 51 percentattended exclusively on a full-time, full-year basis.Thus, what one might call “traditional” fresh-men—i.e., those who are dependent studentsattending on a full-time, full-year basis—comprisea minority of the freshmen population, and itmakes sense to look at them separately fromother types of students.

� For these more traditional freshmen (full-time,full-year, dependent students), substantialpercentages of students in all income quintilesattended public two-year institutions in 1999-2000, but not majorities. (Table Five.) Thepercentages were related to income quintile.Higher-income quintiles were more likely toattend most types of four-year institutions, andwithin four-year institutions, were more likelyto attend more selective (PhD-granting)institutions. Over time, several shifts occurred.The lowest-income quintile saw a shift towardpublic two-year institutions. The lower-middle-income quintile saw simultaneous movementsaway from proprietary and non-PhD four-yearinstitutions, toward both public two-year andPhD four-year institutions. A similar situationoccurred in the higher-income quintiles—shiftsaway from less selective four-year schools,toward both public two-year institutions andmore selective, PhD four-year institutions.21

award levels for Pell Grants and other grantprograms, as well as increases in federal loanlimits. Finally, this period saw the creation of thefederal unsubsidized loan program and, morerecently, a substantial expansion of tax benefits forstudents and their families, especially those in themiddle-income brackets.

Structural changes. Some sectors ofpostsecondary education restructured over thisperiod, in terms of closures, consolidations, lossof federal Title IV aid eligibility, or refocusing ofprogrammatic offerings. This frequently occurredin response to competitive pressures and tofederal regulatory policies regarding loan defaultrates. In particular, the proprietary sectorunderwent substantial transformation, withhundreds of trade schools closing after losingtheir eligibility for federal aid programs andthrough consolidations occurring within theindustry. Some for-profit sectors experiencedtremendous growth, however, especially thosethat are degree-granting (Burd 1997; Borrego2001; NCES 2002).

The adoption of new technologies inpostsecondary education had a substantial impacton the industry. New providers emerged tocompete with existing higher education institu-tions in providing distance learning opportunities.In general, these shifts expanded the number ofalternatives available to students.

Finally, the past decade saw a huge increase ininstitutional marketing, as part of efforts to improvethe quality of consumer-based information aboutinstitutional alternatives. This increase in informa-tion is important in that many students and theirfamilies are more aware of the broad array ofpostsecondary options available to them.

context • continued

21 Not all of the differences between years are statistically significant.


27

� On the other hand, large majorities of the most“non-traditional” freshmen were attendingtwo-year/proprietary institutions in 1999-2000.Focusing on the group of first-year undergradu-ates who were both financially independentand attending either part time or part year(Table Six), students from the lower quintileswere more likely to be attending proprietaryschools than students from the higher incomequintiles. Over time, there was a substantialshift in enrollment away from proprietaryschools and toward public two-year institutions(especially among students from the lowerincome quintiles), similar to the trend amongfreshmen as a whole.22

These results suggest that changes in the composi-tion of postsecondary enrollment were occurringover the past decade, with many non-traditionalstudents choosing to enroll in public two-yearinstitutions instead of proprietary schools, andmore traditional freshmen presenting a complicatedpattern related to family income.

Enrollment distribution data in thecontext of the choice processThe previous two sections have suggested thatstudents with certain characteristics are less likelyto be enrolled at certain types of institutions andthat changes in enrollment patterns have occurredover time, suggesting a movement toward lower-priced institutions (especially community colleges)for many groups of students. These patternsalone, however, are not enough to conclude thatdisadvantaged students are constrained by mon-etary barriers to choice. Ideally, the previousanalyses would be conducted with controls foracademic and other factors in order to isolate the

potential effects of monetary factors; but thedataset used is not conducive to such analysis.23

Prior research (using other datasets) suggests thatthe reality is quite complex. For one thing, moststudents seem able to attend what they label astheir “first choice” institution. Among 1988 eighthgraders who enrolled in four-yearinstitutions by 1994, for example, 71 percentindicated that they were able to attend their firstor second choice institution, with no differencesby race/ethnicity, SES, and test quartiles(Sanderson et al. 1996). This again reinforces thenotion that outcomes cannot be evaluated withoutreference to the entire choice process; decisionsto narrow the choice set come at an earlier stage,and so a student’s “first choice” institution wasinfluenced by a number of factors—bothmonetary and non-monetary—long before anactual selection is made.24 Another study (Berknerand Chavez 1997), using the same longitudinaldata, found that among 1988 eighth graders whograduated from high school, low-income studentswere able to attend four-year colleges at the samerate as students from middle-income families ifthey took the steps necessary to become college-qualified and enroll in a four-year institution—take college preparatory classes, entrance exams,and apply to college. In other words, college-qualified, low-income students who wereaccepted for admission to public and privatefour-year institutions were just as likely to enrollas middle- and upper-income students. On theother hand,low SES and minority students are lesslikely to take those steps, and therefore less likelyto enroll in four-year institutions overall.Importantly, even low-income students whowere academically qualified were less likely totake the tests and apply.

22 Unlike the total freshmen population, however, the differences between years are not statistically significant.23 Test scores (SAT and ACT) are available in NPSAS and could be used as a proxy for academic ability, but they are only available for a minority of students. Better measures,

such as high school grades and curricula, can be used in longitudinal datasets such as NELS.24 Lest one think that there is complete consensus on which factors influence the narrowing of institutions, the Chronicle of Higher Education recently pointed to continuing

debate on whether students narrow their aspirations because of academic preparation/awareness issues, or affordability issues.


28

Taken together, the research suggests that at theaggregate level, a cross-sectional examination ofoutcomes supports the idea that unequal educa-tional opportunity exists for students withdifferent income backgrounds, particularly ifchoice is defined as the ability to attend institu-tions other than the lowest-priced institution.From a longitudinal perspective, however, therole of socioeconomic background in the type ofcollege attended remains important, but appears

to diminish in comparison with factors such asacademic preparation and the application pro-cess. In general, students who acquire collegequalifications while still in high school are morelikely to enroll in college, and in four-year institu-tions, than those who do not. Nonetheless, lowSES students are still less likely to obtain thenecessary academic qualifications, and even ifthey do, they are less likely to take the stepsnecessary for college enrollment.25

25 Most of this research was conducted for students making an immediate transition from high school to college. The conclusions may be quite different for non-traditional students,especially those who have been out of high school for some time.

29

I f the financial aid system as a whole (orfederal need-based aid specifically) is promotingchoice, one might expect to find no substantial

differences by family income in the enrollment ratesof qualified students at particular types of institu-tions. The previous chapter, in noting thatlow-income, college-qualified students may beenrolling at four-year (or more selective) institutionsat rates similar to those of their higher-incomecounterparts, suggested that monetary factors maynot be posing an obstacle for these students.

However, the analysis is not conclusive; it does notreveal whether or not the absence of financial aidwould present a problem or influence studentbehavior, or if current amounts of financial aid, atminimum, are necessary to maintain the currentlevel of choice for these particular students. And, itdoes not reveal anything about whether an influ-ence on choice is even a possibility, given the wayfinancial aid is allocated.

The next level of outcomes analysis examines theeffectiveness of financial aid policy by looking directlyat patterns of financial aid—especially federal need-based aid—to see if the aid is being targeted in a waythat would provide choice (all else being equal). Thisapproach assumes that if aid were promoting choice,it would be allocated in certain identifiable ways.Outcomes data may not be able to measure directlyan influence on student behavior, but outcomes can

measure whether aid is being allocated the way itwas intended.

In particular, federal need-based aid is allocatedthrough the federal need analysis framework,thereby serving the goal of choice directly andindirectly. In the most direct sense, federal aidawards are based on a student’s “need” for financialassistance, defined as the difference between thetotal price of attending an institution and theamount students and their families can contribute(the EFC, which represents ability to pay).26 Thecomponents of the formula mean that the calcula-tion of need can reflect low levels of familyresources, the high price of attending a particularinstitution, or both. Federal need-based financial aidis therefore expected to foster student accessbecause calculated need increases as family abilityto pay decreases; it is expected to encouragestudent choice because need increases as astudent’s price of attendance increases.

The other aspect of the federal need analysisframework is the expectation that federal need-based aid (especially grants) provides a de factoplatform for the neediest students, a platform onwhich other types of aid may be awarded. Federalneed-based aid frequently does not cover all of astudent’s need, as calculated through the needanalysis formula. Financial aid from othersources—including state governments, colleges’

a level deeperexamination of financial aid patterns

26 According to the federal need analysis methodology (FM), the EFC—an attempt to objectively measure families’ ability to pay—is calculated using parent and studentincome, assets, and family size, among other factors. It is helpful to think about need analysis in two stages: first, the EFC is calculated; then, awards of aid are made, basedon the difference between the price of attendance for a student and the EFC (the amount of financial need). One caveat: certain financial assistance (such as veteransbenefits) must be subtracted from the price of attendance before federal aid is awarded. This analysis assumes that the EFC represents family ability to pay somewhataccurately; however, that is also an issue of some debate.


30

internal funds, private donors, and federal non-need-based aid such as unsubsidizedloans—therefore acts in combination with federalneed-based aid in meeting need. Much (but notall) of the non-federal aid is explicitly based onneed, although the calculation may differ.27 Over-all, aid from other sources also promotes studentchoice, whether in combination with federalneed-based aid, or not.

Again, the following sections are based on previousresearch models updated with the most recent dataavailable. Each of these sections examines the target-ing of financial aid from a slightly different perspective;together, these perspectives help explain how varioustypes of financial aid are allocated.

System-wide targeting of financial aidIn evaluating whether financial aid patterns are in linewith the goal of promoting choice, the framework ofthe federal need analysis system argues for a view ofthe financing system as a whole. If financial aid overallis promoting choice, one would expect to see thatstudents attending higher-priced institutions aremore likely to receive aid, all other things beingequal. Of students with the same level of familyincome, those attending more expensive institutionsshould be more likely to receive financial aid, and toreceive higher average amounts.

The Congressional Budget Office (1991) tested thishypothesis by comparing how student aid is sup-posed to be awarded with how it actually is

allocated. CBO examined the patterns of full-time,dependent undergraduates receiving financial aidfrom all sources, as well as the average amountsreceived, in Fall 1986, by family income back-ground, at different types of institutions.28 Theyfound that the percentages of students receiving aidwere higher at more expensive institutions (asmeasured by the type of institution), and thatstudents most likely to receive aid were in thelowest family income category attending the mostexpensive institutions.29 Similar patterns existed forfederal and state aid as for all aid, but receipt ofinstitutional aid depended more on the price ofattendance than on the income level. In terms ofthe type of aid received, grant aid patterns weresimilar to patterns for overall aid, while the per-centages of students receiving loans dependedmore on the price of attendance and less on familyincome background than did the percentages ofstudents receiving any aid. This suggested to CBOthat loans (and institutional aid) were more impor-tant in expanding choice than access, in comparisonwith grants (and federal and state aid).

Using the CBO methodology, this analysis exam-ined more recent patterns of financial aid by type ofinstitution and family income background in 1999-2000 to see if aid is being targeted to promotechoice. (Table Seven.) Controlling for some of theother factors that might affect aid distribution(attendance intensity and dependency status)involved considering the percentages of full-time,full-year dependent undergraduates receiving aid,and the average amounts received. Patterns of aid

27 States and institutions often use the federal need analysis methodology, but may use a separate formula to calculate eligibility for aid awards. Aid packaging models generallystart with outside funds first (federal and state), then follow with institutional funds last if needed; with higher income students who do not qualify for need-based aid, however,institutional aid may be the first in the package. This process leaves room for much discretion, especially regarding the use of institutional grants and federal loans.

28 CBO (1991) examined only full-time, dependent undergraduates, to control for other factors that might influence financial aid patterns (awards of financial aid take intoaccount such factors as attendance patterns and dependency status). Using full-time, full-year, dependent undergraduates allows a relatively homogeneous group with lessvariation in prices within institutional types. Although independent, part-time, and part-year students are a sizable and growing proportion of undergraduates, it is moreproblematic to use this framework for this group of students. There is wide variation in terms of tuition and fees and financial aid received, as well as different attendanceand employment patterns, which would be difficult to take into account for the analysis.

29 The CBO analysis assumed that public two-year institutions are the least expensive institutions, followed by public four-year, proprietary, and finally private, not-for-profitfour-year institutions as the most expensive. This replication uses the same premise, i.e., that institutional type can be used as a proxy for price levels. In fact, in 1999-2000, tuition and fees for undergraduate students ranged from $1,338 at public two-year institutions, to $3,349 at public four-year institutions, to $14,588 at privatefour-year institutions (NCES 2002).


31

Tabl

e Sev

en:

Perc

enta

ges

of fu

ll-tim

e, fu

ll-ye

ar d

epen

dent

und

ergr

adua

tes

rece

ivin

g ai

d an

d av

erag

e am

ount

s re

ceiv

ed, b

y in

com

equ

intil

e an

d ty

pe o

f ins

titut

ion,

199

9-20

00

PER

CEN

T R

EC

EIV

ING

AV

ER

AG

E A

MO

UN

T

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tA

ny

aid

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

77

.1%

59

.8%

39

.7%

32

.2%

26

.7%

$4

,10

7$

3,2

90

$2

,40

5$

2,5

31

—Pub

lic 4

-yea

r8

8.1

%8

1.0

%7

3.7

%6

3.9

%5

0.9

%$

7,8

72

$7

,07

0$

6,1

13

$5

,91

9$

5,7

01

Priva

te for

-pro

fit9

2.4

%8

9.3

%9

1.8

%8

9.5

%6

8.1

%$

9,1

67

$9

,99

2$

9,6

11

$1

0,6

26

—Priva

te n

ot-fo

r-pr

ofit

4-y

ear

95

.8%

88

.3%

87

.4%

87

.3%

68

.9%

$1

4,1

12

$1

6,1

16

$1

5,4

03

$1

4,2

75

$1

2,5

71

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tN

eed-b

ase

d a

idquin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

74

.2%

52

.3%

20

.2%

8.1

%4

.6%

$3

,62

9$

2,5

63

$1

,52

1—

—Pub

lic 4

-yea

r8

2.5

%6

8.3

%5

0.7

%2

6.7

%1

0.5

%$

6,5

58

$5

,36

0$

3,8

87

$3

,04

1$

2,6

91

Priv a

te for

-pro

fit9

2.1

%8

8.1

%7

7.4

%6

7.6

%2

3.5

%$

5,7

34

$5

,19

3$

3,9

58

$3

,82

2—

Priv a

te n

ot-fo

r-pr

ofit

4-y

ear

92

.8%

82

.1%

77

.1%

66

.1%

36

.4%

$1

0,9

81

$1

0,7

38

$9

,87

8$

7,9

28

$6

,27

2

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tG

rants

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

75

.1%

53

.6%

31

.8%

23

.4%

18

.9%

$3

,31

6$

2,2

53

$1

,23

7$

1,8

67

—Pub

lic 4

-yea

r8

5.2

%6

8.5

%4

9.8

%3

5.7

%3

0.6

%$

4,7

20

$3

,97

7$

3,0

67

$3

,21

8$

3,3

32

Priv a

te for

-pro

fit8

4.9

%6

5.7

%3

8.9

%2

9.5

%1

7.8

%$

3,9

28

$3

,26

3—

——

Priv a

te n

ot-fo

r-pr

ofit

4-y

ear

93

.7%

83

.6%

77

.9%

79

.1%

56

.1%

$9

,26

7$

9,9

99

$9

,67

9$

8,4

09

$7

,85

4

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tLo

ans

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

15

.6%

17

.8%

15

.3%

11

.5%

4.6

%$

2,8

70

$3

,65

1—

——

Pub

lic 4

-yea

r5

3.2

%5

4.0

%5

0.8

%4

0.6

%2

6.9

%$

4,3

02

$4

,31

1$

4,3

70

$4

,66

5$

4,6

58

Priv a

te for

-pro

fit6

9.0

%8

0.2

%8

4.0

%8

2.8

%5

7.4

%$

5,2

46

$5

,78

5$

5,2

66

$6

,11

0—

Priv a

te n

ot-fo

r-pr

ofit

4-y

ear

64

.5%

69

.5%

71

.5%

66

.3%

39

.2%

$5

,95

1$

6,2

52

$5

,84

0$

6,0

80

$6

,16

9


32

Tabl

e Sev

en:

Perc

enta

ges

of fu

ll-tim

e, fu

ll-ye

ar, d

epen

dent

und

ergr

adua

tes

rece

ivin

g ai

d an

d av

erag

e am

ount

s r e

ceiv

ed, b

y in

com

equ

intil

e an

d ty

pe o

f in

stitu

tion,

1999-2

000

PER

CEN

T R

EC

EIV

ING

AV

ER

AG

E A

MO

UN

T

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tW

ork

study

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

9.1

%5

.3%

3.0

%0

.6%

0.0

%—

——

—Pub

lic 4

-yea

r1

7.1

%1

5.9

%1

0.9

%4

.4%

2.7

%$

1,7

53

$1

,57

5$

1,8

33

$1

,46

2$

2,0

12

Priva

te for

-pro

fit2

.6%

1.9

%6

.5%

0.0

%0

.0%

——

——

—Priva

te n

ot-fo

r-pr

ofit

4-y

ear

35

.7%

39

.1%

34

.6%

29

.2%

15

.1%

$1

,44

3$

1,5

39

$1

,67

1$

1,6

19

$1

,71

0

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tFed

eral A

idquin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

70

.2%

45

.0%

16

.3%

12

.1%

5.7

%$

3,2

73

$2

,59

3$

2,4

92

——

Pub

lic 4

-yea

r8

0.2

%6

5.6

%5

2.7

%4

0.5

%2

7.1

%$

5,7

17

$5

,24

1$

5,0

12

$5

,52

7$

5,8

89

Priv a

te for

-pro

fit9

2.1

%8

9.0

%8

5.4

%8

2.0

%5

6.0

%$

7,9

38

$7

,92

3$

7,9

07

$9

,82

9—

Priv a

te n

ot-fo

r-pr

ofit

4-y

ear

92

.1%

78

.6%

72

.5%

65

.7%

39

.9%

$7

,14

0$

6,8

50

$6

,71

5$

6,5

62

$8

,09

7

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tS

tate

Aid

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

35

.5%

27

.9%

12

.2%

7.3

%4

.8%

$1

,33

0$

1,4

57

——

—Pub

lic 4

-yea

r3

9.4

%3

3.5

%2

2.1

%9

.6%

7.0

%$

2,1

90

$2

,26

3$

1,8

94

$2

,03

6$

1,9

57

Priv a

te for

-pro

fit2

6.1

%2

3.8

%1

5.7

%1

0.0

%2

.3%

$2

,65

6—

——

—Priv a

te n

ot-fo

r-pr

ofit

4-y

ear

43

.9%

40

.9%

33

.2%

20

.0%

8.3

%$

2,8

94

$3

,17

4$

2,7

26

$2

,60

6$

2,7

50

Low

est

Low

erU

pper

Hig

hes

tLo

wes

tLo

wer

Upper

Hig

hes

tIn

stit

uti

onal A

idquin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

quin

tile

mid

dle

Mid

dle

mid

dle

quin

tile

Pub

lic 2

-yea

r or

less

24

.7%

16

.2%

12

.1%

8.6

%1

1.1

%$

77

5—

——

—Pub

lic 4

-yea

r3

1.9

%3

2.2

%2

6.0

%2

1.4

%1

8.7

%$

2,6

98

$2

,52

7$

3,0

28

$3

,11

0$

3,4

15

Priv a

te for

-pro

fit1

3.0

%2

6.7

%2

4.7

%1

5.1

%1

2.8

%—

——

——

Priv a

te n

ot-fo

r-pr

ofit

4-y

ear

61

.5%

70

.3%

72

.1%

73

.7%

52

.1%

$7

,28

8$

8,2

65

$8

,46

2$

7,6

44

$7

,29

1

— m

eans

not

eno

ugh

sam

ple

size

to c

alcu

late

.No

te: A

vera

ge a

mou

nts

are

for t

hose

rece

ivin

g ai

d. P

LUS

loan

s ar

e ex

clud

ed fr

om th

e an

alys

is. F

eder

al a

id a

lso

excl

udes

ver

tera

ns/D

OD b

enef

its. •

Sou

rce:

NCES

199

9-20

00, b

ased

on

CBO

1991


33

Financial aid and ability to payTo further evaluate the targeting of financial aid, theCBO study compared net prices and expectedfamily contributions (EFC). Net price is defined asthe price of attendance after subtracting financialaid from all sources, or the actual amount a studentmust pay for higher education. The logic of federalneed analysis (as well as the implied framework forfinancial aid overall) suggests that net prices are“reasonable” if they are less than or equal to theEFC; a student can then attend an institutionwithout paying more than need analysis has deter-mined he and his family can pay. In this formulation,the standard of net price equal to or less than EFCcan be used to determine whether access is beingachieved at a particular type of institution.

But, if the net price exceeds the EFC—as it doesfor many groups of students—then a certainamount of “unmet need” exists.30 CBO argued thatsome unmet need will generally exist becauseotherwise, students would have strong incentivesto choose the most expensive school that wouldadmit them and colleges would have incentives toraise their tuitions. Therefore, evaluating thetargeting of financial aid toward choice wouldinvolve determining the extent to which net pricecan rise relative to EFC at schools of choice and stillbe reasonable. How much unmet need is reason-able for students who want to choose moreexpensive institutions? Again, this question is bestaddressed within the context of the financingsystem as a whole, of which federal need-based aidawards are one component.

The CBO analysis worked within the framework oftwo definitions of choice, one based upon level ofinstitution (two-year/four-year) and one basedupon control (public/private). Where public two-year institutions were the institutions of basicaccess, then four-year colleges became the institu-tions of choice; where public four-year institutions

from all sources were used to assess the directionof the financing system as a whole, and patterns offederal aid specifically were explored to determineif federal aid is being targeted as intended.

� The highest percentage of students receivingany aid was students in the lowest incomequintile attending the most expensive type ofinstitution. In general, higher proportions ofstudents received aid at more expensiveinstitutions, in all income quintiles. For ex-ample, 40 percent of middle-income studentsattending public two-year institutions receivedaid, compared to 87 percent of those attendingprivate, not-for-profit, four-year institutions.Also, average amounts of aid received tendedto be higher at more expensive institutions.

� Regarding the source of aid, the percentages ofstudents receiving federal and state aid weresimilar to the pattern of overall aid. For institu-tional aid, substantially higher proportions ofstudents received aid at private not-for-profitfour-year institutions than at other institutionaltypes, especially in the middle-income quintiles.This suggests that from the student perspective,a primary function of institutional aid may be topromote choice of private not-for-profit four-year institutions, which tend to be the mostexpensive institutions.

� Regarding the type of aid, the percentages ofstudents receiving grant aid tended to followthe pattern of overall aid, although the percent-ages appeared to be tied more to familyincome than to type of institution. Other typesof aid differed, however. The highest percent-ages of students receiving work-study aid, byfar, were at private not-for-profit four-yearinstitutions. The percentages of studentsreceiving loan aid appeared to be tied muchmore to the type of institution (price) than tofamily income, suggesting that loans werepromoting choice more so than grants.

30 The concept of unmet need is an artificial construct, based on estimates of financial need. Perhaps a better way of thinking about it is the portion of estimated need that hasnot been met through the financial aid process.


34

were the institutions of basic access, then privatefour-year institutions became the institutions ofchoice. The study found that students in mostfamily income categories (with the exception of thelower-middle-income category) who chose a publicfour-year institution over a public two-year institu-tion had an average net price that was below theaverage EFC, suggesting some level of choice forstudents. On the other hand, choosing privateinstitutions resulted in net prices that were higherthan EFCs, on average.31

Using data for 1999-2000, the following was foundfor full-time, full-year dependent undergraduates(Table Eight):

� Within each income quintile, the ratio of netprice to EFC tends to increase with moreexpensive institutions. At the same time, theratio of net price to EFC at schools of choice(under both definitions) is higher for students

in the lower family income quintiles. Together,these tendencies suggest that the choice ofmore expensive institutions is a relativelymore costly one for students from low-income quintiles than for students fromhigh-income quintiles.

� The choice of more expensive institutions formiddle and upper income students involvedreasonable increases in the ratio of net priceto EFC, under both definitions of choice. Forexample, for middle-income students, theratio increased from 94 percent at public two-year institutions to 104 percent at publicfour-year institutions, and to 150 percent ifthey chose to attend private not-for-profitfour-year institutions. And for the two highestincome categories, a choice of even the mostexpensive type of institution still resulted innet prices that were less than EFCs (i.e.,ratios below 100 percent).

Net price = price of attendance less all aid

Lowest Lower Upper Highest quintile middle Middle middle quintile

Public 2-year or less 532% 218% 94% 61% 30%Public 4-year 356% 189% 104% 70% 40%Private for-profit 987% 259% 90% 81% 58%Private not-for-profit 4-year 569% 274% 150% 98% 73%

Adjusted net price = price of attendance less grant aid and 40 percent of loan aid

Lowest Lower Upper Highest quintile middle Middle middle quintile


Net price exceeds EFC =Net price is lower than EFC =Note: Analysis includes students who did not receive aid. Price of attendance is the attendance-adjusted student budget, and includes tuition, room, board, andother expenses. Loan aid excludes PLUS loans. • Source: NCES 1999-2000, based on CBO 1991

Table Eight: Ratio of net price to expected family contribution (EFC) for full-time, full-year,dependent undergraduates, by income quintile and type of institution, 1999-2000

31 When only loan and work-study subsidies were counted in their definition of financial aid, net prices exceeded EFCs even at public two-year institutions for most income categories.


35

� On the other hand, students in the lowerincome quintiles had net prices that far ex-ceeded their EFCs, even at public two-yearinstitutions. For example, the ratio of net priceto EFC faced by the lowest-income studentsattending public two-year institutions was over500 percent. Since this pattern is true across alltypes of institutions, this type of analysis is lessuseful in examining choice for economicallydisadvantaged students. However, it is interest-ing to note that for the lowest income quintiles,choosing a public four-year institution over apublic two-year institution actually lowers theaverage ratio of net price to EFC.32

� The earlier study found that students in manyfamily income categories could choose a publicfour-year institution over a public two-yearinstitution and still have an average net pricethat was below the average EFC. In 1999-2000,this was true only for the highest two incomequintiles. This suggests that the ability of thefinancial aid system to enable the choice offour-year over two-year institutions may havebeen eroded over time.

These results present a mixed picture of whetherfinancial aid is being awarded in a way that reflects thegoal of promoting choice. To an extent, the conclu-sions depend upon the definition of choice, as well asthe income background of the student. If two-yearinstitutions are the standard for meeting access, thenone could argue that the current financial aid system isworking well if the costs of choosing a public four-yearinstitution over a public two-year institution arereasonable. The data show that for students in themiddle and upper income quintiles, making this typeof choice involves relatively small increases in the ratioof net price to EFC; for the lowest income quintiles,making this choice may involve a decrease in the ratio,but for these students net prices remain far aboveEFCs on average. If public four-year institutions

represent basic access, then financial aid is beingallocated in line with the goal of choice if the costs ofchoosing proprietary or private four-year institutionsare reasonable. In this case, the data again show thatmaking this choice involves relatively small increases inthe ratio of net price to EFC for middle and upperincome students. For students in the lower incomequintiles, however, it would lead to substantialincreases in the ratio, within the context of unmetneed at all types of institutions.

Proportion of federal aiddedicated to choiceThe previous analysis examined the targeting offinancial aid within the context of the financingsystem as a whole. It is also possible to specificallylook at the targeting of federal need-based aid, thefoundation of the federal need analysis framework(Table Nine).33 The results show similar patterns, butwith a few differences. In general, the ratios of netprice to EFC are higher, reflecting the fact that onlyfederal need-based aid was used to reduce the priceof attendance. Students in the middle- and upper-middle-income quintiles at more types of institutionsface net prices that exceed average EFCs. In addi-tion, the choice of private not-for-profit four-yearinstitutions (over both public four-year institutions,and public two-year institutions) involves substantialincreases in the ratio of net price to EFC for almostall income groups, suggesting that non-federal aid(such as institutional aid) may play a significant role inenabling choice at these institutions.

CBO researchers took the analysis a step furtherby trying to estimate the proportion of federal aidthat went toward promoting choice (as opposed toaccess). To do this, they examined the differencebetween the (face value) amount of aid awarded atthe school of choice and the amount of aid thestudent would have received attending an average-

32 This effect virtually disappears when using adjusted aid, suggesting that loans are the primary vehicle reducing unmet need for lower-income students choosing publicfour-year institutions.

33 Federal need-based aid includes Pell Grants, campus-based aid such as SEOG and Perkins loans, and Stafford subsidized loans.


36

Lowest Lower Upper HighestAll quintile middle Middle middle quintile

SCHOOL OF ACCESS: PUBLIC 2-YEARUsing all aid

Public 4-year 53% 46% 63% 85% 92% 100%Private for-profit 75% 51% 70% 91% 98% 100%Private not-for-profit 4-year 73% 61% 75% 93% 98% 100%

Using adjusted aidPublic 4-year 34% 34% 50% 85% 88% 84%Private for-profit 67% 42% 61% 91% 97% 96%Private not-for-profit 4-year 57% 50% 63% 93% 96% 97%

SCHOOL OF ACCESS: PUBLIC 4-YEARUsing all aid

Private for-profit 47% 8% 21% 38% 71% 74%Private not-for-profit 4-year 43% 28% 34% 53% 70% 82%

Using adjusted aidPrivate for-profit 50% 12% 22% 42% 72% 72%Private not-for-profit 4-year 34% 25% 26% 51% 68% 78%

Note: Each ratio is calculated as: (amount of aid at school of choice - amount of aid at school of access)/amount of aid at school of choice. Analysis is basedon average amounts of aid for all students, including those who did not receive aid. Adjusted aid is federal grants plus 40 percent of federal need-based loans(Perkins and Stafford subsidized). •Source: NCES 1999-2000, based on CBO 1991

Table Ten: Estimated proportion of federal need-based aid used to pay for schools of choicefor full-time, full-year, dependent undergraduates, by school of access, type of institution, andincome quintile, 1999-2000

Net price = price of attendance less all federal need-based aid

Lowest Lower Upper Highestquintile middle Middle middle quintile


Adjusted net price = price of attendance less federal grant aid and 40 percent of federal need-based loan aidLowest Lower Upper Highestquintile middle Middle middle quintile


Net price exceeds EFC =Net price is lower than EFC =Note: Analysis includes students who did not receive aid. Price of attendance is the attendance-adjusted student budget, and includes tuition, room, board, andother expenses. Loan aid excludes PLUS loans. Federal need-based loans include Perkins and Stafford subsidized loans.Source: NCES 1999-2000, based on CBO 1991

Table Nine: Ratio of net price to expected family contribution (EFC) for full-time, full-year,dependent undergraduates, by income quintile and type of institution, 1999-2000Using only federal need-based aid in calculation of net price


37

priced school of access. The following analysisestimates the proportions of federal need-based aidfor 1999-2000.34 Once again, the implicationsdepend on the definition of the basic institution ofaccess, public two-year institutions or public four-year institutions (Table Ten):

� When public two-year institutions are consid-ered the basic institutions of access, about half(53 percent) at public four-year institutions, andabout three-quarters at private institutions, of allstudents' federal need-based aid went topromote choice. This varied by family income,with higher percentages of aid going to promotechoice the higher the income quintile. In thehighest income quintiles, virtually all of federalneed-based aid went to promote choice.

� When public four-year institutions are the basicinstitutions of access, slightly less than half offederal need-based aid for all students went topromote choice (43 percent of aid for privatenot-for-profit institutions, 47 percent of aid forproprietary institutions). This varied accordingto income quintile, with about three-quarters offederal aid promoting choice for the highestincome quintiles (compared to less than one-third for the lowest income quintile).

The patterns of federal aid suggest that a substan-tial portion of need-based aid is being allocated asit was intended with regard to the goal of promot-ing choice. It is also clear that the extent to whichfederal aid is consistent with the goal of choicedepends on how choice is defined.

34 At the time of the original CBO report, the major non-need-based federal programs (PLUS loans for parents and Stafford unsubsidized loans) were not a major part of the equation.

39

T his report describes the various concep-tions of college choice and how they havechanged over time. In addition, it highlights

what is known about both the process of choice andthe factors that influence each step of the process, aswell as the outcomes of college choice decisions.

Both the process and the outcomes are useful topolicymakers in evaluating public policy. Under-standing how students make enrollment decisionshelps policymakers decide how and where tointervene to improve progress toward the goal of“choice,” however it is defined. Examining wherestudents are actually enrolled helps policymakersdetermine if policy interventions are “working”and how much more intervention is needed toreach the goal of choice—which again dependson the specific definition of choice. Another partof the outcomes approach involves exploringwhether financial aid (a specific policy) is beingtargeted in the way it was intended to support aspecific goal.

The analyses in this report raise several issues:

Choice is not easy to define.The specific meaning of college choice depends onone’s perspective, and usually must be describedwith reference to something else (the definition ofaccess, for example). In addition, the term “choice”is used frequently as a codeword for specific policygoals. In this sense, choice may mean not just theenabling of options for students, but also mayconnote issues of institutional quality and the desireto equalize tuition levels between sectors. Sharpen-ing the definition of choice and clarifying the role of

conclusionsspecific forms of financial aid in promoting choice willenable better targeted public policy.

Should choice be a goal of federal policy?And if so, is the way choice is currently expressedthrough need analysis the appropriate vehicle? Theexisting framework defines financial “need” in a waythat is sensitive to the price of attendance, in order tosupport the goal of choice. However, this means thatstudents can be defined as “needy” based upon theprice of the institution they choose, rather than as aresult of disadvantaged economic circumstances.What is the appropriate balance?

Choice is not just about price differentials.Many non-monetary factors influence students’behavior at various points in the choice process,including academic ability and preparation, institu-tional characteristics, and other factors. Thissuggests that early intervention programs andother pre-college activities, as well as K-12curricula reform efforts, are important aspects ofthe choice equation.

Data on outcomes suggest that the choices ofcertain groups of students are constrained.Low-income students, as well as students with“non-traditional” characteristics, appear to be lesslikely to enroll at four-year and private institutionsthan their higher income, traditional counterparts.In addition, changes in enrollment patterns haveoccurred over time, suggesting a movementtoward lower-priced institutions (especially com-munity colleges) for many groups of students.Outcomes data do say something about the “sta-tus” of choice and fit the perception that many


40

students are choosing lower-priced institutions dueto increasing tuition levels. However, from out-comes-based analysis alone, it is not clear whetherthe cause of these differential rates of enrollment ismonetary, academic, some other factor, or acombination of factors.

To judge the effectiveness of public policy, andwhen to effectively intervene, it is importantto look at a combination of analyses of stu-dent behavior and outcomes.Altogether, the research suggests that publicpolicies enhance, but do not equalize, choice forcertain groups of students. A high proportion ofstudents appear to enroll at their first or secondchoice institutions, but enrollment patternssuggest that choice may be constrained earlier inthe process by a combination of monetary andnon-monetary factors. Disadvantaged studentswho take all of the necessary steps and arecollege qualified enroll in four-year institutions atsimilar rates as advantaged students; however,disadvantaged students are less likely to take therequisite steps, even those with high levels ofacademic ability.

The data present a mixed picture of whetherfinancial aid is being awarded in a way thatreflects the goal of promoting choice.The extent to which aid patterns are consistentwith the goal of choice depends not only on howchoice is defined, but also on the income back-ground of the student.

The context of the higher education systemis important to any evaluation of thestatus of choice.Public policies change over time and the composi-tion of students attending postsecondaryeducation institutions and the structure of theindustry itself shift in response to competitivepressures and other forces. Given the ongoingpatterns of tuition increases—with growing gapsbetween two-year and four-year tuitions, andpublic and private tuitions—choice might beeroded even without change in federal policy.Further, given the increases in institutional alterna-tives and the availability of information, somewell-informed groups of students may haveexperienced an increase in college choice at thesame time that other groups faced a decline.

41

ACSFA. See Advisory Committee on Student FinancialAssistance.

Advisory Committee on Student Financial Assistance(ACSFA). 2001. Access Denied: Restoring the Nation'sCommitment to Equal Educational Opportunity.Washington, DC: ACSFA, February.

Akerhielm et al. 1998. Factors Related to CollegeEnrollment. Prepared for the Office of the Under-Secretary, U.S. Department of Education. Princeton,NJ: MathTech, Inc.

Berkner, Lutz, and Lisa Chavez. 1997. Access toPostsecondary Education for the 1992 High SchoolGraduates. NCES 98-105. Washington, DC: Govern-ment Printing Office, October.

Borrego, Anne Marie. 2001. “A Wave of ConsolidationHits For-Profit Higher Education.” Chronicle ofHigher Education. August 10.

Breneman, David W., and Chester E. Finn, Jr. 1978. “AnUncertain Future.” In David W. Breneman andChester E. Finn, Jr., eds. Public Policy and PrivateHigher Education. Washington, DC: The BrookingsInstitution.

Burd, Stephen. 1997. “Is the Law on Student-LoanDefaults Too Tough or Not Tough Enough?”Chronicle of Higher Education. September 19.

Cabrera, Alberto F., and Steven M. La Nasa. 2000.“Three Critical Tasks America' s Disadvantaged Facein their Path to College.” In Alberto F. Cabrera andSteven M. La Nasa, eds. Understanding the CollegeChoice of Disadvantaged Students. New Directionsfor Institutional Research No. 107. San Francisco:Jossey-Bass Publishers, 23-29.

Carnegie Commission on Higher Education. 1973. HigherEducation: Who Pays? Who Benefits? Who Should Pay?New York, NY: McGraw Hill.

Carnevale, Anthony P., and Stephen J. Rose. 1998.Education for What? The New Office Economy.Executive Summary. Princeton: Educational TestingService.

CBO. See U.S. Congressional Budget Office.

Choy, Susan P., and Cecilia Ottinger. 1998. Choosing aPostsecondary Institution. NCES 98-080. Washington,DC: Government Printing Office, October.

College Board. 2000. Trends in Student Aid 2000. NewYork: College Entrance Examination Board.

Cost Commission. See National Commission on the Costof Higher Education.

Gladieux, Lawrence E., and Thomas R. Wolanin. 1978.“Federal Politics.” In David W. Breneman and ChesterE. Finn, Jr., eds. Public Policy and Private HigherEducation. Washington, DC: The Brookings Institution.

Hearn, James C. 1991. “Academic and NonacademicInfluences on the College Destinations of 1980High School Graduates.” Sociology of Education 64(July): 158-171.

________. 1988. “Determinants of Postsecondary Educa-tion Attendance: Some Implications of AlternativeSpecifications of Enrollment.” Educational Evaluationand Policy Analysis 10, no. 2 (Summer): 171-185.

Heller, Donald E. 1997. “Student Price Response inHigher Education: An Update to Leslie andBrinkman.” The Journal of Higher Education 68, no. 6(November/December): 624-659.

references


42

Hossler, Don, and Karen S. Gallagher. 1987. “StudyingStudent College Choice: A Three-Phase Model andthe Implications for Policymakers.” College andUniversity 2 (Spring): 207-221.

Hoxby, Caroline M. 1998. “The Return to Attending aMore Selective College: 1960 to the Present.”Unpublished paper, Harvard University, Boston, MA.

Institute for Higher Education Policy. 1999. State ofDiffusion: Defining Student Aid in an Era of MultiplePurposes. Prepared for the New Millennium Projecton Higher Education Costs, Pricing, and Productiv-ity. Washington, DC: Institute for Higher EducationPolicy, The Education Resources Institute, and theFord Foundation, August.

________. 1999b. What is Opportunity? Defining,Operationalizing, and Measuring the Goal ofPostsecondary Education Opportunity. Washington,DC: Institute for Higher Education Policy, TheEducation Resources Institute, and the Council forOpportunity in Education.

Lee, John B. 1999. “How Do Students and Families Payfor College?” In Jacqueline E. King, ed. Financing aCollege Education: How It Works, How It' s Changing.Phoenix: Oryx Press.

Manski, Charles F., and David A. Wise. 1983. CollegeChoice in America. Cambridge, MA: HarvardUniversity Press.

McDonough, Patricia M. 1997. Choosing Colleges: HowSocial Class and Schools Structure Opportunity. Albany:State University of New York Press.

McPherson, Michael S., and Morton Owen Schapiro.1999. “Reinforcing Stratification in American HigherEducation: Some Disturbing Trends.” Paper writtenfor the Macalester Forum on Higher EducationConference, Diversity and Stratification in AmericanHigher Education, Macalester College, June.

________. 1998. The Student Aid Game. Princeton:Princeton University Press.

NCES. See U.S. Department of Education. NationalCenter for Education Statistics.

Paulsen, Michael B. 1990. College Choice: UnderstandingStudent Enrollment Behavior. ASHE-ERIC HigherEducation Report No. 6. Washington, DC: TheGeorge Washington University, School of Educationand Human Development.

National Commission on the Cost of Higher Education.1998. Straight Talk About College Costs and Prices.Washington, DC: Oryx Press, January.

Rouse, Cecilia Elena. 1994. “What to Do after HighSchool: The Two-Year versus Four-Year CollegeEnrollment Decision.” In Ronald G. Ehrenberg, ed.Choices and Consequences: Contemporary Policy Issuesin Education. Ithaca: ILR Press.

Sanderson, Allen et al. 1996. National Education Longitu-dinal Study 1988-1994 Descriptive Summary Report:With an Essay on Access and Choice in PostsecondaryEducation. NCES 96-175. Washington, DC: Govern-ment Printing Office, May.

St. John, Edward P., Michael B. Paulsen, and Johnny B.Starkey. 1996. “The Nexus Between CollegeChoice and Persistence.” Research in HigherEducation 37, no. 2 (April): 175-220.

U.S. Congressional Budget Office (CBO). 1991.Student Aid and the Cost of Postsecondary Educa-tion. Washington, DC: Government PrintingOffice, January.

U.S. Department of Education. National Center forEducation Statistics (NCES). 2002. Digest of Educa-tion Statistics 2001. NCES 2002-025. Washington,DC: Government Printing Office, February.

________. 2002b. Condition of Education 2002. NCES2002-130. Washington, DC: Government PrintingOffice.

________. 1999-2000. National Postsecondary StudentAid Study, 1999-2000 (NPSAS:2000), Undergradu-ate Data Analysis System.

________. 1989-90. National Postsecondary Student AidStudy, 1989-90 (NPSAS:90), Undergraduate DataAnalysis System.


43

Wolanin, Thomas R. 2001. Rhetoric and Reality: Effectsand Consequences of the HOPE Scholarship. The NewMillennium Project on Higher Education Costs,Pricing, and Productivity. Washington, DC: Institutefor Higher Education Policy, April.

________. 1998. “Pell Grants: A 25-Year History.” InLawrence E. Gladieux, Bart Astor, and Watson ScottSwail, eds. Memory Reason Imagination: A QuarterCentury of Pell Grants. New York: College EntranceExamination Board.

Expanding Student Options in Higher Education · THE POLICY OF CHOICE: Expanding Student Options in Higher Education viii In 1999-2000, 62 percent of all first-year under-graduates

Documents