Roberto M. Ifill and Michael S. McPherson

V o l u m e 5 • N u m b e r 2 • J u l y 2 0 0 4

NEW AGENDA SERIES™

When SavingMeans Losing:

Weighingthe Benefits of

College-savingsPlans

Acknowledgments

The authors are grateful to Bridget Terry Long, Peter Orszag and Morty Schapiro for their comments onearly drafts of this report. We would also like to thank Lumina Foundation for its financial support ofthis work. The opinions expressed in this monograph are those of the authors and do not necessarilyreflect policies or positions of Lumina Foundation for Education, its officers or members of its board ofdirectors.

Executive summary1

Introduction3

A brief history of 529 plans5

Mechanics of the need-based aid system8

The impact of saving on needs assessment11

Summary and conclusions14

References17

Endnotes18

Table of contents○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

1

Executive summary○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

S ince the early 1980s, college tuitionshave soared, and state and federalgovernments have sought new ways tohelp students and families meet the costsof attendance. Annual state and federal

appropriations to traditional student aid programshave more than doubled in the past two decades.In addition, the federal government created theHope Scholarship and Lifetime Learning tax creditprograms, and state governments created prepaidtuition plans and college-savings plans. These statesavings plans, called “529 plans,” are used by manyfamilies to help pay for college. However, becauseof the way funds invested in the different plans aretreated by traditional financial aid programs,participation in them can affect eligibility forscholarships, grants and loans. Depending on thedetails, an increase in savings in a 529 plan mayresult in a dollar-for-dollar decrease in eligibilityfor need-based grant aid or it may result in nodecrease at all. This paper describes the ways inwhich this might happen for different groups ofstudents at different types of institutions.

One of the things families should consider indeciding whether to invest in a prepaid tuitionplan, a 529 plan, some general financial savingsplan (such as a mutual fund), or a savings accountin a son’s or daughter’s name is the interactionbetween the level of accumulated assets and the

need-based financial aid system. In general, themore assets a family accumulates, the less aid it willbe eligible for. The precise relationships varyenormously, however,depending on the savingsinstrument chosen, thefamily’s level of incomeand wealth, the cost ofattendance at the chosencollege, and the details ofthat institution’s financialaid policies.

How need analysissystems assess a family’sability to pay for college isa critical element in theinteraction betweensavings plans andtraditional financial aidprograms. In general, thegreater the family’s annualearnings and accumulated wealth and the fewerfamily members financially dependent upon thoseresources, the more that family can reasonably beexpected to pay toward college costs. Thisassessed amount is called the Expected FamilyContribution (EFC). But the EFC is only one partof the financial need equation. To determineeligibility for need-based aid, the EFC is subtracted

An increase insavings in a 529plan may resultin a dollar-for-dollar decreasein eligibility forneed-basedgrant aid.

http://www.luminafoundation.org

from the student expense budget (the charges fortuition and fees, room and board, books andsupplies, and other expenses students incur). Theremaining amount is the student’s financial need.When the EFC is greater than the student expensebudget, the student is determined to have no need(for financial aid purposes) and will be ineligiblefor aid from need-based programs. When studentexpense budgets are very high (to attend eliteprivate colleges, for example), families withrelatively high EFCs could still demonstratefinancial need. Therefore, one must look at boththe student budgets and family financial resourcesto analyze the effects of savings plans on eligibilityfor aid.

This report analyzes the effects on financial aidof a $100 increase in assets in each of four familyincome levels and three costs of attendance. Theincome levels (low, middle, upper-middle andhigh) correspond to approximate family incomesof less than $30,000; $50,000-$60,000; $80,000;and $125,000 or more. Low-priced colleges aretypically community colleges. Medium-pricedinstitutions are “flagship” public universities forstudents paying in-state tuition and typically livingaway from home. High-priced institutions are themore expensive privates and some out-of-statepublics.

For families expecting to receive need-basedaid, prepaid tuition plans are disadvantageousbecause the need analysis system views this moneyas a resource readily available for college payment.Just as when a student receives an outsidescholarship, the presumption is that prepaid tuitiondollars reduce need dollar for dollar.

There are two important conclusions to bedrawn from the analyses in this report. First,although high-income families have no financialaid disincentive to save for college, lower-incomefamilies do. By saving, especially by choosing aprepaid tuition plan to save, lower-income familiesmay reduce their eligibility for need-based aid.Second, unless they are low-income families whoexpect their children to attend communitycolleges, families have little reason to worry thattheir savings decisions will affect their eligibilityfor need-based aid. However, their choices ofsavings vehicles can influence the amounts bywhich their need-based aid may be reduced.

The specific characteristics of a savingsinstrument can have a disproportionate impact onits treatment under the need analysis system.There is little relevant difference in the motivationof a parent who puts money into a savings plan, aprepaid tuition plan or a savings account in a son’sor daughter’s name. Likewise, there is little relevantdifference among these programs in terms of thefederal interest in encouraging families to save forhigher education. Yet families choosing these plansmay find their need-based aid reduced modestly,substantially or dramatically, depending on whichinstrument they choose. There is clearly goodreason to better align the outcomes of thesecollege-savings plans with the intentions of thosewho created and support them. This is a challengethat is likely to be addressed in the next reauthori-zation of the Higher Education Act.

2

3

Introduction○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

These funds havebecomeincreasinglypopular amongfamilies seeking toshelter themselvesfrom the risk ofrising tuitions andtax liability.

I n 2001, Congress passed the largest packageof tax cuts since 1981. Most of the publicitycentered on the slashing of marginal rates,but there were numerous specialized tax cutsand enhanced deductions as well. Of

particular interest to this study are the enhance-ments to the tax benefits affecting Qualified StateTuition Plans originally introduced in 1997. Theseplans, also known as 529 plans (in reference to theunderlying section of the Internal Revenue Codethat applies to them), offer families the opportu-nity to pay all or a substantial part of the costs ofcollege, either by “prepaying” tuition or byaccumulating wealth over time that can be appliedto college expenses. In recent years, these fundshave become increasingly popular among familiesseeking to shelter themselves from the risk ofrising tuitions and tax liability. State-sponsoredinstitutions and, more recently, some selectiveprivate institutions have promoted prepaid tuitionplans, and all 50 states have established college-savings schemes or have plans to do so in the nearfuture.

The most straightforward of the new taxprovisions is directed at college-savings funds.Prior to the 2001 tax changes, a parent or otherbenefactor could invest in such funds and not paytaxes on its accumulated proceeds until applyingthem to tuition for one or more beneficiaries. At

that time, income drawn from the fund foreducational purposes would be treated as earnedby the student and taxed accordingly. Under thenew provision, thisincome is not subject tofederal tax if it is devotedto payment of collegetuition and fees. Manystates also provide taxbenefits related to theseplans, either by makingcontributions deductibleor by waiving taxation ofearnings, or both. Theseincentives are likely toinduce people to shiftexisting assets into thesesavings plans, to directnew savings into suchplans, and perhaps toincrease the total amountthey save for college. Theencouragement to greater saving induced by thehigher return may be offset in part by the fact thatpeople can achieve their savings targets with fewerdollars because of the higher return.

However, neither the original creation nor therecent expanded benefits for 529 plans hasoccurred in a vacuum. Both have been accompa-

4

nied by other changes in tax law or financial aidpolicies that might also affect saving. The marketfor college-savings or “tuition futures” is arelatively new one, and the array of financingoptions for families has grown rapidly, oftencausing confusion. Changes in governmental andinstitutional financial aid strategies and optionshave also contributed to this trend. Federal policiesoriginally marketed as a means to provide“universal access” to college via programs such asthe Hope Scholarship and Lifetime Learning TaxCredits (although aimed principally at providingtax relief to middle-income parents), add morefeatures to a tangled landscape.

Of particular importance is the complexinteraction between programs that provide savingsincentives and need-based student aid programs.There has long been concern that need-based aiddiscourages families from saving, since the greaterwealth resulting from saving increases familyability to pay and therefore reduces the size ofneed-based aid awards (other things equal). Theactual impact of this negative incentive on familysaving behavior remains controversial (Case and

McPherson, 1986; Feldstein, 1995; Edlin, 1997),yet the growth and proliferation of the newsavings vehicles have sharpened these concerns.Depending on the details, an increase in savings ina tax-preferred vehicle may result in a dollar-for-dollar decrease in the size of a need-based grant,or it may result in no decrease at all. The resultsdepend on the particular savings vehicle chosen,the level of a family’s income and wealth, and theprice of the college in question.

In this paper, we aim to sort out the interac-tions between need-based financial aid and familyinvestments in 529 plans by working through anumber of cases covering alternative familycircumstances and college alternatives. We hopesuch a guide will assist three important groups:

1. Families, as they consider their savingsoptions.

2. Colleges, as they consider how to treatthese accumulated savings in their ownfinancial aid policies.

3. Policy-makers, as they consider both fed-eral student aid policy and federal taxpolicy.1

5

A brief history of 529 plans○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

With anincreasing numberof Americansaspiring tocollegeattendance,rising prices havespurred concernabout access andaffordability.

R ising college costs have been a source offamily anxiety and political and policyconcern since the 1980s. With tuition

increases exceeding inflation at public and privatecolleges and universities, and with an increasingnumber of Americans aspiring to college atten-dance, rising prices have spurred concern aboutaccess and affordability.

Responding to this concern, state governmentshave provided novel ways for families to prepare inadvance to pay for college. In the late 1980s,Wyoming, Michigan and Florida began anexperiment in college financing. They offeredparents the opportunity to purchase up to fouryears of college credits at current prices to beredeemed when a designated recipient enrolled atan eligible institution; these instruments werecalled prepaid tuition plans. These plans were atfirst restricted to in-state residents and onlyapplicable to a small number of state-supportedinstitutions. In essence, the states offered to sharefinancial risk with tuition payers — gainingpredictable enrollments by offering stable prices.(See Government Accounting Office [1995] for ahistorical overview on which we have drawn here.)

Initially, relatively few households chose thisoption, perhaps because the upfront expense wasprohibitively high or because the restrictions onwhere the credits would be applied were too

inflexible. Of course, there was the risk thatstudents would not wish to attend the designatedinstitutions, or that they would not gain admissionor persist to graduation.Nevertheless, over the nexteight years, the originalthree states were joined bysix others, all offeringprepaid tuition plans.

Starting with Kentuckyin 1990, a number of statesalso offered a more flexiblemeans for financing college.Under what were called“college-savings plans,”households could contrib-ute to an investmentportfolio, much like an IRA,up to specified annuallimits. Their investmentswould be managed by astate agency or a financialfirm contracted to the statefor a modest fee, and thegross proceeds could thenbe applied to college-related expenses when thebeneficiary enrolled at an eligible institution.Unlike the prepaid plans, which were akin to“defined benefit” pension instruments, college-

6

By now, all but afew states have

some form ofQSTP (also

known as“529 plans”).

savings plans operated as “defined contribution”instruments.

A number of developments in the late 1990squickly led most states to adopt both types ofsavings plans. First, the prolonged bull market ofthis period encouraged growing numbers ofhouseholds to invest in mutual funds, IRAs, 401ksand similar plans; the increasing acceptance of thisform of saving removed a conceptual barrier fromstates contemplating the creation of such plans.Not coincidentally, the improving fiscal strengthof the states and higher education as a wholereduced the financial risk for savings plan sponsorsand encouraged financial firms to compete for thebusiness. Second, a surge in personal incomeincreased the fiscal capacity of families withcollege-age children. On the one hand, this mademore disposable income available for saving; onthe other hand, it decreased eligibility for need-based aid.

In addition to these developments, a number ofpublic policy changes at the state and federal levelshelped spread these plans to nearly all states by

the beginning of 2001.As college-savings plansbecame more popular,states removedresidency restrictionson eligibility. Nowcitizens of any statecould invest in most ofthese plans, and theirproceeds could beapplied to out-of-stateinstitutions, both publicand private. Some statesallowed for “rebalanc-

ing” investments, permitting holders to adjust theproportion invested in equities, bonds or otherassets.

In 1996 and 1997, federal tax policy took noteof these plans, designating them as Qualified StateTuition Plans (QSTPs) under Section 529 of theInternal Revenue Code and providing favorable taxtreatment for the income deriving from these

plans. In other words, although QSTPs wereconsidered parental assets, accrued income wouldbe treated as deriving from the recipient; usually,this meant a substantial reduction in the marginaltax rate (from as high as 39.5 percent to 15 percentor less). Further, investors could avoid the $10,000annual cap on untaxed gifts by contributing up to$50,000 to the fund, providing that no furthercontributions to the particular beneficiary wouldbe made for five years. Fund owners could switchintended beneficiaries without penalty, as long asthese were also family members (excludingcousins).

By now, all but a few states have some form ofQSTP (also known as “529 plans”). Purchasers canchoose from among hundreds of different plansacross the nation and use them to pay costs atalmost any accredited institution. A number ofprivate institutions developed a proposal to createa Tuition Plan Consortium to offer prepaid plansfor its members. In the Economic Growth and TaxRelief Reconciliation Act of 2001, Congressprovided even more favorable treatment for 529plans. Income from savings plans was now to beexempt from federal income tax, households couldtransfer assets from one QSTP to another withoutchanging beneficiaries or incurring penalties, andcousins were now allowed to be designatedbeneficiaries. Investments in 529 plans have grownsubstantially, with total invested assets increasingfrom $8.6 billion in 2000 to $34.6 billion by June2003 (Schmidt, 2003).

In recent years, the relative popularity ofsavings plans and prepaid tuition plans hasreversed abruptly. In 2000, 71 percent of assetswere in prepaid plans, while in 2003 75 percent ofthe assets were in savings plans (Schmidt, 2003).The principal reasons for this shift were the steepdrop in the stock market and unusually rapidincreases in public tuition, a combination thatthreatened the viability of the prepaid plans. Stateofficials who were optimistic during the stockmarket boom that returns on the funds would beadequate to finance rising tuitions lost confidencewhen fortunes reversed. Several states with prepaid

7

plans have suspended their programs or closedthem to new participants, and states that lack suchplans are no longer introducing them. A secondaryreason for the movement away from prepaid plansmay be their treatment in needs analysis systems,as discussed below.

There were other significant developments infederal policy. In the same tax reform package thatloosened restrictions on 529 plans, Congressintroduced a new tax deduction option for collegeexpenses, an alternative to 1997’s Hope Credit;now households could choose either to take theHope tax credit of up to $1,500 or to deduct up to$3,000 from taxable income. The legislation alsoendeavored to make Coverdell Education SavingsAccounts more attractive by increasing themaximum annual contribution to $2,000. Contri-butions to these accounts are not tax-deductible,but earnings are exempt from income tax ifwithdrawals are used to pay for qualified educationexpenses. From 2002 on, qualifying expenses havebeen extended to include elementary andsecondary school expenses. The Coverdell taxbenefits are phased out for joint filers withincomes above $190,000 and single filers withincomes above $95,000 (Ma and Fore, 2002). InJanuary 2004, Congress further increased theattractiveness of Coverdell plans by permittingthem to be treated as parental, rather than student,assets in the federal needs analysis methodology.This change substantially reduced the financial aid“taxing rate” on these plans.

Policies and practices governing financial aidalso have undergone momentous change in recentyears. Loans have grown steadily as a proportionof financial aid. The purchasing power of Pellgrants has fluctuated, but despite rapid growthbetween 1997 and 2002, it remains substantiallylower than in the early years of the program.Colleges continued to adjust the ways theycalculated expected family contributions, changesreflected in the financial aid methodologiesdeveloped by the College Board and the federalgovernment. Colleges have become more strategicin using financial aid to meet institutional goals,

through greater use of merit aid and throughadjusting the loan-grant mix in financial aidpackages as tools to attract the students they want.

Our brief historical review shows that, even astuition and fee increases continue to outpaceinflation, families have a larger array of options forfinancing college. However, that array has becomemore complex; it’s nosurprise that, in the past fewyears, college financialadvisers have established arobust cottage industry forthose families that canafford them.

One of the thingsfamilies must consider indeciding whether to investresources in a particularsavings instrument is theinteraction between thelevel of accumulated assetsand the need-based financial aid system. Ingeneral, the more assets a family accumulates, theless aid it will be eligible for. The precise relation-ships vary enormously, however, depending on thesavings instrument chosen, the family’s level ofincome and wealth, the cost of attendance at thechosen college, and the details of that institution’sfinancial aid policies. A further complication, ofcourse, is that policies regarding financial aid maychange between the time a family begins accumu-lating assets and the time their children are readyfor college.

No simple analysis can capture the full range ofvariation in these possibilities. In the remainingsections of this paper, we focus on a simplifiedanalysis, but one which we believe captures someof the central dimensions of variation. We feel itallows us to draw some conclusions about who ismost likely to benefit from the various savingsoptions recently made available through federallegislation.

In general, themore assets afamilyaccumulates, theless aid it will beeligible for.

8

○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

The formula forcalculating the

EFC assumes thata family’s abilityto pay increases

more thanproportionally

with theparents’ income.

A s a prelude to our analysis of interactionsbetween the aid system and savingsplans, it is necessary to review the basic

operation of a need-based financial aid system. Atthe core of such a system lies a method for

assessing a family’s abilityto pay for college. At thistime, the calculationsused in this assessment atmost institutions are builton one of two approaches— the Federal Methodol-ogy, legislated byCongress, and theInstitutional Methodol-ogy, established by theCollege Board andsubject to various federalguidelines. While the twomethods differ in anumber of details, theyare broadly similar inapproach, with thelargest difference being

that the Federal Methodology ignores the equityin a family’s home as a resource that might bedrawn on in paying for college.2

For simplicity, our discussion here and theanalysis below will focus on the Federal Methodol-

ogy, but nothing essential would change if westudied the Institutional Methodology. Eithermethodology draws on information about afamily’s income, assets, debts, expenses and futureobligations to come up with a number called the“Expected Family Contribution” (EFC) — anestimate of what the family can be expected to paytoward college expenses. The formula forcalculating this number assumes that a family’sability to pay increases more than proportionallywith the parents’ income. It also assumes that, forany given level of income, a family with moreaccumulated assets has more ability to pay. Assetsand income of the student also are assessed forexpected college contributions, and at a substan-tially higher rate than parental income and assets,presumably because the student is the directbeneficiary of the education.

The difference between the cost of attendanceat a particular college and the EFC is the family’s“demonstrated need” (which is of course set to zeroif the EFC is greater than the cost of attendance ata particular institution). How (and if) that need ismet is a complicated matter. In general the collegeadmitting the student serves as the “packager” ofaid resources. These resources include grants(which don’t need to be repaid), loans (which doneed to be repaid, though often with some subsidyfrom the federal government), and work to be

Mechanics of the need-based aid system

9

done by the student during school terms. Depend-ing on the wealth of the family and the state ofresidence, need-based grants may come from thefederal and/or state governments, to a smallerextent from outside philanthropies, and from thecollege itself. (Grants from the institution are ineffect “discounts” against the cost of attendance.)Loans for the most part are backed by the federalgovernment, although some institutions alsoprovide loans directly. For the most part, work ispaid for by the school, often with substantial helpfrom federal funds.

One key element of the financial aid packagefor most lower-income students is set by formula.That element is the Pell grant, where awardeligibility and size are calculated according to aformula that, like the calculation of the EFC, issensitive to family income and assets. Some statesadd a formula-driven need-based grant on top ofthe Pell grant for students with need. Most otheraid is “packaged” by the institution, with the aim(in a pure need-based system) of filling the gapbetween the EFC and the cost of attendance.Typically a student will be presented with such apackage of federal, state and institutional grantsalong with recommended levels of loans andearnings from work (at jobs supplied by theschool) to meet need.

Schools have considerable discretion inadjusting the elements of the aid package otherthan those dictated by federal and state formulas.They may choose not to meet full need; they mayoffer differing mixes of loan, grant and work todifferent students with the same need (presumablybased on how much they want to “land” thatstudent); and they may offer grants unrelated toneed to students they want to attract. The mainconstraint colleges face in packaging aid is that thetotal amount of aid they provide to a student whoreceives federal aid may not exceed that student’sneed (as calculated in the Federal Methodology).

This discretion on the part of schools greatlycomplicates efforts to assess the impact of families’college-savings decisions on a student’s aid award.While more saving will clearly raise the EFC (by

differing amounts depending on how savings areheld, as we will see below), the bottom-line impacton the cost of college to the family will depend onthe school’s packaging philosophy. Thus, at oneextreme lies Princeton, which meets all need withgrants. At Princeton, every dollar of added EFCmeans a dollar less of grants. If you are sure (1)that your daughter will be admitted to Princeton,(2) that she will choose to attend, and (3) that youwill be eligible for need-based aid at Princeton atthe time your daughter attends, you can be surethat added saving will resultin a smaller grant for yourdaughter to attendPrinceton. Note, though,that virtually no one is sureof these three things; also,even if you are sure, it maystill pay to save because, aswe will see, your EFC andtherefore your grant isgenerally reduced by farless than a dollar for everyadditional dollar you save.

At another extrememight lie a school whichdevotes no resources toneed-based grants.Whatever money goes intofinancial aid is used formerit awards for high-ability students. If your sonwill attend such a school, and if your income ishigh enough that he will not be eligible for a need-based Pell grant, your son’s grant award is likely tobe unaffected by increases in your EFC.3 At most,the higher EFC might result in your son’s beingasked to take a larger loan or to work more duringthe term.

It is obviously impossible to capture all of thepotential variations among colleges in their aidpolicies in the illustrative cases developed below.Instead, we focus on one key variable: theresponsiveness of the level of need (as assessed bythe Federal Methodology) to changes in the level

Schools haveconsiderablediscretion inadjusting theelements of theaid packageother than thosedictated byfederal and stateformulas.

10

of saving. This change in the need level may, asthe extreme examples above indicate, result inequal changes in grant, in a mix of changes ingrant, loan and work, or in no change in grant atall. It is our sense that in most cases, students whoare eligible for need-based grants will receive alower grant award if their families have a higherEFC, although probably typically not on a dollar-for-dollar basis. Nonetheless, we think the simplest“shorthand” for assessing the effect of an increasein the level of saving is the size of the decrease inthe level of need, for those students who are in facteligible for grant aid.

We note, finally, that there is another level ofcomplexity we ignored entirely in assessing therelationship between federally subsidized savingsprograms and financial aid. This is the matter ofindependent students.4 Adult students who areindependent of their parents have their ability topay assessed according to a methodology that is,so to speak, “tacked on” to the methodology that isused for “dependent” students. This presents a hostof additional complexities regarding the relation-ship between saving and financial aid — complexi-ties that, although interesting in their own right,we put aside for this analysis.

11

O ur aim here is to present a simplifiedquantitative analysis of the responsive-ness of need-based aid awards to

variation in accumulated assets held in variousforms. We focus on families with dependentstudents. We consider families in four incomegroups: low, middle, upper-middle and high; andwe consider three price ranges for colleges: low,medium and high.

Calculations based on these categories arereported in Table 1 on Page 16. The focus is oncases in which a family (if eligible) will receiveneed-based aid from the institution the studentattends in addition to any other aid, such as federalor state grants. Because the institutional award isgenerally the “last dollar in” — that is, the finalfactor in the financial aid formula — it is thisaward that is sensitive to changes in the family’sasset position. (It is not uncommon for studentsfrom low-income families to attend low-pricedinstitutions that award little or no aid. For suchcases, it is variation of the Pell award with changesin the family’s asset position that is relevant totheir decisions. Because the Pell grant awardformulas are similar to those for other need-basedaid, the analysis for those cases is essentiallysimilar to what we report.)

Note that needs analysis methodology protectsa certain level of assets from the calculation of

The impact of saving on needs assessment○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Because theinstitutional awardis generally the“last dollar in,”it is this awardthat is sensitive tochanges in thefamily’sasset position.

parents’ ability to contribute, recognizing the needto plan for emergencies and for retirement. Thecalculations here thus only apply to increments inassets above the protected amount. For a two-parent family with theolder parent aged 50, theallowance is $47,900(Federal MethodologyNeeds Analysis Tables,2004). Families in thelowest income class areunlikely to have enoughassets to be subject tofinancial aid “taxation” atthe margin.

Table 1 is based onthe following scenario:Consider a family with ason or daughter who hasjust completed highschool and is poised toattend college. We ask:How would a $100increment in the family’s accumulated financialassets affect the amount of need they would beshown to have under the federal needs analysismethodology? The answer to this questiondepends on how the marginal $100 in assets isheld, and we consider four cases: (1) a 529 savings

12

plan, (2) a 529 prepaid tuition plan, (3) a savingsaccount in the student’s name, and (4) a non-education-related savings vehicle (such as mutualfund).5 In the table, we refer to this last item as“General Financial Assets.” We show separately theimpact in the first year of college, and the impactover four years. For comparative purposes, we alsoshow how a $100 increase in after-tax income

would affect the needanalysis, but readersshould be aware thatthis is an “apples andoranges” comparison —we are reporting theimpact on assessed needfrom changes in theflow of income and inthe stock of assets.

How should weinterpret the fourincome categories?Roughly, “low income”relates to families whichare (a) eligible for Pellgrants and (b) face thelowest marginal taxingrate in the needsanalysis system (22percent). Depending on

family circumstances, this typically corresponds tofamilies earning under about $30,000 per year.Middle-income families are (a) typically ineligiblefor Pell and (b) face an intermediate taxing rate inthe needs analysis system (34 percent). This mightcorrespond to an annual family income of $50,000-$60,000. Upper-middle-income families are beingtaxed at the highest rate in the needs analysis (47percent), and might have incomes around $80,000per year. High-income families have income highenough to eliminate need at even the high-pricedinstitutions, and therefore are out of the needsanalysis system altogether. Depending again onfamily circumstances as well as the price of theinstitution attended, this might correspond toincomes of $125,000 or more.

We interpret the three categories of institu-tions in the following way: Low-priced institutionsare typically community colleges, with very lowtuition and many students who live at home.Medium-priced institutions are “flagship” publicuniversities for students paying in-state tuition andtypically living away from home. Some lower-priced private colleges also fall in this category.High-priced institutions are the more expensiveprivates and some out-of-state publics.

To see how Table 1 works, consider the case ofa middle-income family whose child is to enroll ina medium-priced institution. A $100 increase inthe parents’ after-tax income will result in a $34decrease in the family’s assessed need. A $100increase in the amount of money the family holdsin a prepaid tuition account will result in a $100decrease in measured need (we will return to thisdramatic result later). An extra $100 in a 529savings plan (or Coverdell plan) will result in a$4.10 decrease in measured need in the first year,the same impact as a $100 increase in holdings in anon-education-related savings vehicle. A $100increase in holdings in an account in the student’sname will result in a $35 decrease in measuredneed. For the cases of 529 savings plans, studentsavings accounts and non-tax preferred vehicles,we also report the cumulative effects over fouryears. Thus, with general financial assets, 2.66percent of the $100 increment is “taxed” by theneeds analysis system in the first year, leaving$93.73, which is again “taxed” at 2.66 percent, andso on. After four years, about $90 of the original$100 is left.6

Reviewing Table 1 as a whole, there are threebroad results to underscore. First, as familyincomes rise, families in effect “top out” of theneeds analysis system. This happens quickly atlow-priced institutions such as communitycolleges, where even middle-income families areunlikely to be eligible for aid. Obviously, in thissituation, increments in income or in assets haveno impact on the level of measured need, which isalready zero. High-income families are in thissituation at all categories of institutions. (Note that

By saving —and especially by

choosing the“wrong” savings

instrument —low-incomefamilies canreduce their

eligibility forneed-based aid.

13

some upper-middle income families may be toppedout of the needs analysis system at medium-pricedinstitutions, although here we show that they arestill affected.) There are two important conclusionsto be drawn here. First, though high-incomefamilies have no disincentive to save for college,lower-income families do: By saving — andespecially by choosing the “wrong” savingsinstrument — they can reduce their eligibility forneed-based aid. Second, unless they are low-income families who expect their children toattend community colleges, they have little reasonto worry that their savings decisions will affecttheir eligibility for need-based aid.7

For families expecting to receive need-basedaid, the choice of savings instruments makes a bigdifference. The dramatic response of needsanalysis to a prepaid plan comes about because theneeds analysis system views the money madeavailable by the plan as a resource directlyavailable for college payment. Just as when astudent receives an outside scholarship, thepresumption is that tuition dollars that have beenprepaid reduce need dollar for dollar.8 This impliesthat all of the money accumulated in a prepaidplan will be taken before any need-based aid isprovided. This aid treatment of course implies thatfamilies expecting to be eligible for need-based aidshould not invest in prepaid plans.9

Accounts in the student’s name suffer from asimilar but less severe difficulty. The assetsaccumulated in a student’s name are treated asowned by the student. Thus assets accumulated onbehalf of a student are viewed as student assets,and the needs analysis system presumes that alarge portion of such assets (35 percent) should bemade available for college, a much higher fractionthan applies to parental assets. For parents’ assets,the annual percentage ranges from 2.7 percent to5.6 percent, depending on the family’s income.10

Third, we note that the impact of assetaccumulation on measured need rises with incomeuntil income is so high that the family no longerhas need. Thus a low-income family that managesto put funds into a 529 savings plan will find that,

over four years, the needs analysis system expectsonly about 10 percent of those parental assets tobe drawn down for college expenses; that is, thefamily’s cumulative four-year need is reduced byabout $10 for every added$100 they have in their 529funds at the time of entry tocollege. (Families are ofcourse free to draw down asmuch as they want, but overfour years their need willonly be reduced by about10 percent of the incremen-tal asset value.11) An upper-middle-income family willfind that, at the margin, itsaccumulated savings have much more impact onneed over four years. Such a family will find itsneed reduced by more than 20 percent of theincremental asset value.

These impacts are high enough to affect ratesof return significantly. As a simple example,consider a family that puts aside a sum of $20,00010 years in advance of a child’s entering college. Ifthat sum grows at a continuously compoundedafter-tax rate of 5 percent annually, it will be worth$32,974 after 10 years. If that value is reduced by10 percent (corresponding to a low-incomefamily), the annual rate of return on the investmentdrops from 5 percent to 3.9 percent. If the value isreduced by 20 percent (corresponding to an upper-middle income family), the return drops to 2.8percent. Notice, however, that these implicationsare not specific to saving in 529 plans — theyapply to all additions to financial assets for familiessubject to needs analysis. As noted earlier, there iscontroversy about the degree to which suchdifferences in returns affect decisions about savingsrates, but these are surely substantial differences. Itis important to note that high-income families,who are not subject to the needs analysis system,suffer no similar discouragement to saving forcollege.

Familiesexpecting to beeligible for need-based aid shouldnot invest inprepaid plans.

14

There is clearlygood reason to

better alignoutcomes with

intentions.

○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Summary and conclusions

O ur aim here has been to illuminate,through examination of illustrativecases, the interactions between the

need-based student aid system and the savingsvehicles subsidized by the federal government toencourage saving for college. It is perhaps notsurprising that the interactions between thesesystems are sometimes awkward.

One obvious difficulty is that the specificcharacteristics of a savings instrument can have adisproportionate impact on its treatment under the

needs analysis system.There is little relevantdifference in themotivation of a parentwho puts money into asavings plan, a prepaidtuition plan or a savingsaccount in the student’sname. Nor is theremuch relevant differ-ence among theseprograms from the

standpoint of the federal interest in subsidizinghigher education saving. The tax benefits of thesethree types of instruments are likewise similar.12 Yeta family that has pursued one of these plans andlater finds a son or daughter enrolled at a schoolthat grants need-based aid may find their aid

reduced modestly, substantially or dramatically,depending on which instrument they chose. Thereis clearly good reason to better align outcomeswith intentions here — and, indeed, there is goodreason to question whether there is a soundrationale for providing federal tax subsidy for all ofthe different instruments.13 As noted above, this isa challenge that is likely to be addressed in thenext reauthorization of the Higher Education Act.

The complexity and proliferation of federallyfavored savings instruments present additionaldisadvantages. First, such complexity increasescitizens’ tax-preparation costs — costs thatgenerally benefit accountants and tax preparers.Second, since affluent citizens are more able toafford such tax-planning services, they stand tobenefit disproportionately.

Under current arrangements, it is clear thatfamilies should steer clear of prepaid 529 plans andshould not provide assets as gifts to students unlessthey are confident either that the financial aidtreatment will change before their children areready for college or that the family’s wealth willdisqualify them for need-based aid at the collegestheir children will attend. Since, under current law,Coverdell accounts provide their full tax benefitsonly to joint filers with incomes less than $190,000(and single filers with incomes less than $95,000[Ma, 2003]), only a subset of high-income people

15

will be interested in these accounts. All high-income people can take advantage of 529prepayment plans without risking a financial aidpenalty.

The fact that it is the highest-income familieswho are least likely to face a financial aid penaltyfor investing in federally subsidized savingsprograms — simply because they are outside theneed-based aid system altogether — raises anotherissue, as many may view this inequality in benefitsas improper in itself. Apart from the equityquestion, however, the aim of savings subsidies ispresumably to encourage more saving for collegethan would otherwise occur. But it is the highest-income people, those who are outside the need-based financial aid system, who are least likely tohave their savings decisions influenced by taxsubsidies for college saving. The reason for this isthat high-income people are much more likelythan others to have substantial accumulations of

assets independent of college-savings incentives.Such individuals can realize the tax benefits ofparticular savings instruments simply by reallocat-ing existing assets into tax-preferred instruments. Ifthe goal, then, is to use federal tax subsidies toincrease the amount of college savings amongfamilies, focusing those subsidies on families ofrelatively limited means is consistent with that goal.14

These issues about the implications of college-savings plans for both governmental and institu-tional policy clearly deserve more attention, andwe intend to pursue them further in future work.There is also a need for empirical study of thebehavioral effects of college-saving incentives —work that is, unfortunately, very hard to do well.15

We hope that this contribution has helped toclarify some of the basic relationships betweensavings incentives and the need-based aid systemand, in that way, has helped set the stage for moreproductive discussion and for further work.

16

Table 1. Impact on financial aid of a $100 increase in assetholding or income, by family income and cost of attendance

Cost of attendanceFamily income Low price Medium price High price

One Four One Four One Fouryear years year years year years

Low$100 increase in ... Income -22 NA -22 NA -22 NA

Prepaid plan -100 NA -100 NA -100 NA529 saving -2.66 -10.1 -2.66 -10.1 -2.66 -10.1Student savings -35 -82.1 -35 -82.1 -35 -82.1General financial assets -2.66 -10.1 -2.66 -10.1 -2.66 -10.1

Middle$100 increase in ... Income 0 NA -34 NA -34 NA

Prepaid 0 0 -100 NA -100 NA529 saving 0 0 -4.1 -15.3 -4.1 -15.3Student savings 0 0 -35 -82.1 -35 -82.1General financial assets 0 0 -4.1 -15.3 -4.1 -15.3

Upper-middle$100 increase in ... Income 0 NA -47 NA -47 NA

Prepaid 0 0 -100 NA -100 NA529 saving 0 0 -5.6 -20.7 -5.6 -20.7Student savings 0 0 -35 -82.1 -35 -82.1General financial assets 0 0 -5.6 -20.7 -5.6 -20.7

High$100 increase in ... Income 0 NA 0 NA 0 NA

Prepaid 0 0 0 0 0 0529 saving 0 0 0 0 0 0Student savings 0 0 0 0 0 0General financial assets 0 0 0 0 0 0

Note: We do not report four-year data for changes in income because the relationship displayed is betweenincome in a given year and need reduction in a given year. For other items, the relationship is between the size ofa stock of accumulated saving and the need reduction, first for one year and then accumulated over four years.

See text for full treatment of income groups. In summary, they are:

Low income — Pell-eligible, lowest marginal rate in needs analysisMiddle income — Pell-ineligible, intermediate marginal rate in needs analysisUpper-middle income — Pell-ineligible, top marginal rate in needs analysisHigh income — No need

17

References○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Case, K.E. and M.S. McPherson (1986). Doesneed-based student aid discourage saving forcollege? New York, College Entrance Examina-tion Board: Copies available from CollegeBoard Publications.

Dynarski, Susan (April 2004, unpublished). “TaxPolicy and Education Policy: Collision orCoordination? A Case Study of the 529 andCoverdell Saving Incentives.”

Edlin, A. and A.W. Dick (1997). “The ImplicitTaxes in College Financial Aid.” Journal ofPublic Economics 65(3): 295-322.

“Federal Methodology Needs Analysis Tables.”http://www.fafsa.com/fmtables.htm (Web siteas of May 18, 2004).

Feldstein, M. (1995). “College Scholarship Rulesand Private Saving.” American EconomicReview 85(3): 552-566.

General Accounting Office (1995). CollegeSavings: Information on State Tuition Prepay-ment Programs.

Ma, J. and D. Fore (2002). “Saving for Collegewith 529 Plans and Other Options.” RetirementPlanning (September/October): 43-52.

McPherson, M.S. and M.O. Schapiro (1998).The student aid game: meeting need andrewarding talent in American higher education.Princeton, N.J., Princeton University Press.

Schmidt, P. (2003). Pre-paid Tuition Plans Feelthe Pinch. Chronicle of Higher Education,September 12.

18

1 A recent paper by Dynarski (2004) addressesrelated themes, with an emphasis on the interac-tions between income taxes and the needs analysissystem.

2 Basic information about needs analysis method-ologies is available at the College Board Web site,www.collegeboard.com. A small number of highlyselective colleges (referred to as the “568 group,” alabel that derives from a feature in federal law thatallows these schools to cooperate in assessing theirstudents’ need for assistance) have recentlyadopted a third methodology, a modification ofthe Institutional Methodology. There are somesignificant differences in this methodology that arerelevant to comparison of savings plans — notably,student assets are treated on a par with parents’assets, rather than being assessed much moreheavily. Of course, when parents are saving forcollege, few of them can be confident their son ordaughter will be admitted to or choose to attendone of this limited set of institutions.

3 In some states, your son may receive a smallerstate grant as a result of having a higher EFC.

4 Our analysis also ignores students who attendcollege less than half time; they generally areineligible for federal need-based aid.

5 Note that retirement-savings vehicles such as the401k are not counted in total assets in the financialaid system and therefore do not reduce thecalculated level of need.

6 This calculation makes most sense if the decreasein need is reflected in a decrease in grant, ratherthan in loan. However, the four-year cumulation isa reminder that families must plan on a repeatedimpact of the asset value on the needs analysis overthe student’s time in college.

7 This is true, anyway, of the first two years ofcollege. For families expecting community collegeto be followed by two years of attendance at afour-year institution, the calculation is obviouslydifferent.

8 Individual institutions may use “professionaljudgment” to modify this result either in the caseof outside scholarships or of prepaid tuition.However, institutions are bound by the rule thatsays that no student who receives any federal aidfrom the major (Title IV) programs can receive atotal aid package that exceeds need as measured bythe Federal Methodology. This constrains theability of aid officers to “discount” the impact ofsuch resources on need.

○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Endnotes

19

9 There is active discussion of modifying thisaspect of the federal needs analysis methodologyduring the next reauthorization of the HigherEducation Act.

10 This treatment of Coverdell accounts may alsobe addressed in the upcoming reauthorization.

11 The more rapidly the asset is drawn down, thesmaller is the total impact on need, since the “tax”in later years will be applied to a smaller base.

12 As noted above, there is an income test on thetax benefits of the Coverdell accounts.

13 A prepaid tuition program can really be factoredinto two components: a savings program and atuition insurance program. There is no reason inprinciple why these two elements could not beseparated, with states or groups of schools sellingtuition insurance policies. Whether such insurancepolicies, separated from tax policies, would find amarket at a price that covers costs is questionable.

14 The issues here closely parallel those regardingincentives for retirement saving. For a criticalreview of the literature on the impact of savingsincentives for retirement, see Engen, Gale andScholz (1996). In their piece, they offer thefollowing observation about behavioral responsesto taxation: “... [D]ecisions concerning the timingof economic transactions are the most clearlyresponsive to tax considerations. The next tier ofresponses includes financial and accountingchoices, such as allocating a given amount ofsaving to tax-deferred versus other assets. The leastresponsive category of behavior applies to agents’real decisions, such as the level of saving” (p. 135).

15 The review by Engen, Gale and Scholz (1996) ofattempts to do such empirical work in the relatedcontext of retirement savings incentives under-scores the severe empirical challenges.

About the authors

Roberto M. Ifill is an independent consultant on higher education in Washington, D.C.Previously, he was a Research Fellow at the American Council on Education and served on thesenior staff at Macalester College and at Connecticut College. Before that, Ifill was a programofficer at the Andrew Mellon Foundation and was the dean of freshmen and a member of theeconomics department at Williams College. He holds a doctorate in economics from YaleUniversity.

Michael S. McPherson is president of the Spencer Foundation in Chicago. Previously, he waspresident of Macalester College and before that was the dean of faculty and a member of theeconomics department at Williams College. He has served as a Senior Fellow at the BrookingsInstitution and as a member of the Institute for Advanced Study in Princeton, N.J. He has co-authored several books and a number of articles on higher education finance with MortySchapiro, president of Williams College.

Also available from Lumina Foundation for Education○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○ ○

Expanding College Access:

The Impact of State Finance Strategies

Edward P. St. John, Choong-Geun Chung, Glenda D. Musoba,

Ada B. Simmons, Ontario S. Wooden and Jesse P. Mendez

February 2004

Unintended Consequences of Tuition Discounting

Jerry Sheehan Davis

May 2003

Following the Mobile Student; Can We Develop the

Capacity for a Comprehensive Database to Assess

Student Progression?

Peter T. Ewell, Paula R. Schild and Karen Paulson

April 2003

Meeting the Access Challenge: Indiana’s

Twenty-first Century Scholars Program

Edward P. St. John, Glenda Droogsma Musoba,

Ada B. Simmons and Choong-Geun Chung

August 2002

Unequal Opportunity: Disparities in College

Access Among the 50 States

Samuel M. Kipp III, Derek V. Price and Jill K. Wohlford

January 2002

Hope Works: Student use of

Education Tax Credits

Barbara A. Hoblitzell and Tiffany L. Smith

November 2001

Learning in the Fast Lane: Adult Learners’ Persistence

and Success in Accelerated College Programs

Raymond J. Wlodkowski, Jennifer E. Mauldin

and Sandra W. Gahn

August 2001

Debts and Decisions: Student Loans and Their

Relationship to Graduate School and Career Choice

Donald E. Heller

June 2001

Funding the “Infostructure:” A Guide to Financing

Technology Infrastructure in Higher Education

Jane V. Wellman and Ronald A. Phipps

April 2001

Discounting Toward Disaster: Tuition

Discounting, College Finances, and Enrollments

of Low-Income Undergraduates

Kenneth E. Redd

December 2000

College Affordability: Overlooked Long-Term

Trends and Recent 50-State Patterns

Jerry Sheehan Davis

November 2000

HBCU Graduates:

Employment, Earnings and Success After College

Kenneth E. Redd

August 2000

Student Debt Levels Continue to Rise

Stafford Indebtedness: 1999 Update

Patricia M. Scherschel

June 2000

Presidential Essays: Success Stories —

Strategies that Make a Difference at

Thirteen Independent Colleges and Universities

Allen P. Splete, Editor

March 2000

Are College Students Satisfied?

A National Analysis of Changing Expectations

Lana Low

February 2000

Fifty Years of Innovations in Undergraduate Education:

Change and Stasis in the Pursuit of Quality

Gary H. Quehl, William H. Bergquist and Joseph L. Subbiondo

October 1999

Cost, Price and Public Policy:

Peering into the Higher Education Black Box

William L. Stringer, Alisa F. Cunningham, with

Jamie P. Merisotis, Jane V. Wellman and Colleen T. O’Brien

August 1999

Student Indebtedness:

Are Borrowers Pushing the Limits?

Patricia M. Scherschel

November 1998

It’s All Relative: The Role of Parents in

College Financing and Enrollment

William L. Stringer, Alisa F. Cunningham,

Colleen T. O’Brien and Jamie P. Merisotis

October 1998

Lumina Foundation for Education, a private,independent foundation, strives to help peopleachieve their potential by expanding access and

success in education beyond high school. Throughresearch, grants for innovative programs andcommunication initiatives, Lumina Foundationaddresses issues surrounding financial access andeducational retention and degree or certificateattainment — particularly among underserved studentgroups, including adult learners. Focusing on these areas, the Foundation framesissues and explores new solutions through fact-basedresearch. Because we strive to be a credible andobjective source of information on issues affectinghigher education, Lumina Foundation encouragesoriginal sponsored research. The results of that research,and therefore the content of this and other LuminaFoundation® publications, do not necessarily representthe views of the Foundation or its employees. Believing that published research may have thelongest-term impact on higher education, theFoundation publishes and disseminates articles, researchreports and books. We prefer topics and approachesthat are more practical than theoretical, and whichemphasize pragmatic tools that will assist institutionsand public policy-makers.

Additional information about the Foundation’sresearch program may be obtained from:

Dr. Robert C. DickesonSenior Vice President for Higher EducationPolicy, Research and Evaluationrdickeso@luminafoundation.org

Additional information about Lumina Foundation®

programs and communications may be obtained from:Susan O. ConnerExecutive Vice President andChief Operating Officersconner@luminafoundation.org

Lumina Foundation for EducationP. O. Box 1806Indianapolis, IN 46206-1806800-834-5756www.luminafoundation.org

Lumina Foundation for EducationNew Agenda Series™

Robert C. DickesonExecutive Editor

David S. PowellDirector of Publications

Jean B. RoseCopy Editor

Natasha SwingleyDesign and Production

Printed by SPG Graphics Inc.

Lumina Foundation for EducationEditorial Advisory Board

Deborah G. BonnetSusan O. ConnerJerry S. DavisJill Wohlford

Lumina Foundation for EducationNew Agenda Series™ is published periodically byLumina Foundation for EducationP.O. Box 1806Indianapolis, IN 46206-1806

First-class postage paid at Indianapolis, Ind.Copyright © 2004Lumina Foundation for Education, Inc.All rights reserved.

Send all name and address changes to:Lumina Foundation for EducationAttn: Database Manager3813 E. Winston StreetBloomington, IN 47401-4292nswingley@luminafoundation.org

July 2004

NEW AGENDA SERIES™