Toward a More Complete Understanding of the Role of Financial Aid in Promoting College Enrollment: The Importance of Context

Abstract

Through a review of prior research, this chapter argues that knowledge about how student financial aid programs can best promote college enrollment is incomplete in part because existing research does not devote sufficient attention to the “context” in which these programs operate or the ways that context mediates the effects of aid. The chapter begins by explaining the need to increase educational attainment, the ways inadequate finances limit educational attainment, and the role of financial aid in addressing financial barriers. Then the characteristics of student financial aid programs are described and what is known from existing research about the effects of financial aid on college-related behaviors is summarized. A conceptual model for understanding the ways “context” may influence the relationship between financial aid and college enrollment is proposed. Building on this framework, the chapter concludes by offering questions to guide future research, as well as recommendations for fruitful research strategies.

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Competing demands for public resources and demands for accountability of program effectiveness underscore the need to better understand the role of student financial aid in eliminating financial barriers to college enrollment. Although many researchers have examined the effects of student financial aid on student outcomes, gaps in knowledge persist. Based on their comprehensive review of research for the College Board’s “rethinking student aid” project, Baum and McPherson (2008) concluded that more research is needed, as “many questions remain about how to develop the most effective student aid policies” (p. 6).

Through a review, synthesis, and critique of relevant prior research, this chapter argues that one reason knowledge about how student financial aid programs can best promote college enrollment is incomplete is that existing research does not devote sufficient attention to the “context” in which these programs operate or the ways that context mediates the effects of aid. Potentially relevant aspects of context include characteristics of other available financial aid programs, students and their families, the high school attended, available higher education options, and the broader economic, social, and policy environment. These aspects of context come together to define and delimit the ways financial aid may influence enrollment. Available research also provides insufficient attention to the indirect effects of aid, particularly the ways information and perceptions of financial aid may influence not only college enrollment but also college aspirations and academic preparation (Mundel, 2008). The chapter begins by explaining the need to increase educational attainment, the ways inadequate finances limit educational attainment, and the role of financial aid in addressing these financial barriers. Then the characteristics of student financial aid programs are described and what is known from existing research about the effects of financial aid on college-related behaviors is summarized. A conceptual model for understanding the ways “context” may influence the relationship between financial aid and college enrollment is proposed. Building on this framework, the chapter concludes by offering questions to guide future research, as well as recommendations for fruitful research strategies.

The Importance of Increasing Educational Attainment

Improving college enrollment and degree attainment is critical to ensuring the nation’s continued economic and social prosperity as a growing share of jobs need workers with a college education. The United States Department of Labor’s Bureau of Labor Statistics (BLS, 2003) projects that, on average, employment in occupations that generally require a bachelor’s or associate’s degree will increase faster than employment in other occupations. Jobs generally requiring a college degree comprised 29% of all jobs in 2000 and are expected to account for 42% of projected new job growth during the coming decade. BLS (2003) also predicts that 12 of the 20 fastest growing occupations typically require a bachelor’s or associate’s degree. Analyses of data from the Census Bureau and Current Population Survey show that 69% of white-collar office workers, the largest, fastest-growing, and among the highest-paying categories of employment, had at least some college education in 2001, up from 37% in 1973 (Carnevale & Desrochers, 2003). Moreover, although the number of frontline factory jobs declined by 21 million between 1959 and 2001, the remaining jobs in this sector are increasingly held by workers who have at least some college education (31% in 2001 versus 8% in 1973) (Carnevale & Desrochers, 2003).

Although increasingly important, educational attainment in the United States has fallen behind that of other developed nations (Baum & Ma, 2007; National Center for Public Policy and Higher Education, 2006). The educational attainment of the U.S. adult population has increased over time, as 28% of adults age 25 and older in the U.S. held at least a bachelor’s degree in 2006, up from 26% in 2000 and 21% in 1990 (Baum & Ma, 2007). But, other nations are increasing the educational attainment of their populations at a faster rate (National Center for Public Policy and Higher Education, 2006). Although the quality and focus of available postsecondary education varies across nations, the rate of first-time enrollment into programs that lead to a traditional college degree (i.e., “tertiary-A programs”) is now lower in the United States than in New Zealand, Sweden, Iceland, Finland, Poland, Australia, Norway, and Hungary (Organization for Economic Co-operation and Development [OECD], 2006). The OECD predicts that, because the United States is experiencing both slower enrollment and lower rates of college completion than some other nations, the share of college graduates in OECD nations produced by the United States will decline over the next decade.

Recognizing these trends, in a nationally-televised, prime-time address to a joint session of Congress on February 24, 2009, President Barack Obama presented an ambitious goal: “By 2020, America will once again have the highest proportion of college graduates in the world.” Along the same lines, Lumina Foundation for Education (2009) has a goal of increasing the percentage of young adults in the United States who have attained an associate’s or bachelor’s degree from the current 40% to 60% by 2025. The first goal in the State Higher Education Executive Officers’ (SHEEO, 2008) “national agenda for higher education” is to increase the percentage of young adults who earn at least an associate’s degree to 55% by 2025, which would require producing an additional three million degrees each year between now and 2025 (SHEEO, 2008). The College Board’s Commission on Access, Admissions, and Success in Higher Education (2008) embraces this goal, promising to annually report progress toward achieving it.

One Approach to Increasing Attainment: Eliminate Financial Barriers to Enrollment

One area of opportunity for raising the nation’s production of degree recipients is the population of students who are academically prepared for college but do not attend. The Advisory Committee on Student Financial Assistance (2006) estimates that between 2000 and 2010 1.4–2.4 million students from low- and middle-income families will be academically qualified for college but will not complete a bachelor’s degree because of insufficient financial resources.

Although inadequate academic preparation and other forces also play a role (Adelman, 1999, 2006; Perna, 2005, 2006a), data and research consistently demonstrate that money is a primary deterrent for academically-qualified students who do not enroll in college (Hahn & Price, 2008; Perna, 2006a; St. John, 2003). The positive relationship between college enrollment and family income persists even though postsecondary educational participation rates have increased since the mid-1980s regardless of family income (Baum & Ma, 2007). The magnitude of the gap has fluctuated, but over this period college enrollment rates have been at least 25–30% points lower for high school graduates in the lowest family income quintile than for those in the highest quintile (Baum & Ma, 2007). College choices are also stratified by family income, as students from lower-income families who enroll in college are relatively concentrated in public 2-year and private for-profit institutions and underrepresented at public and private 4-year doctoral-granting universities (Baum & Ma, 2007). In 2003–2004, dependent undergraduates with family incomes below $20,000 represented 26% of all dependent undergraduates attending private for-profit institutions and 16% of those attending public 2-year colleges, but only 10% of dependent undergraduates attending public and private 4-year doctoral-granting institutions (Baum & Ma, 2007).

The positive relationship between family income and college enrollment persists even after controlling for differences in academic achievement. Only 29% of 1988 eighth graders in the highest quartile of math achievement and lowest quartile of socioeconomic status had attained at least a bachelor’s degree within 8 years of their scheduled high school graduation, compared with 74% of those in the highest quartile of math achievement but highest quartile of socioeconomic status (Baum & Ma, 2007). Family income is also positively related to educational attainment among students with the highest SAT scores. Among dependent undergraduates with SAT scores in the top quartile (i.e., 1,100 or higher) who first enrolled full-time in a 4-year college in 1995, only 71% of those with family incomes below $40,000, but 86% of those with family incomes above $70,000, had completed at least a bachelor’s degree by 2001 (Baum & Ma, 2007).

The Role of Financial Aid in Eliminating Financial Barriers to Enrollment

Each year federal and state governments, as well as colleges and universities, foundations and other organizations, invest substantial resources in programs designed to eliminate financial barriers to college enrollment. In 2007–2008 alone, students nationwide received more than $96 billion in federally-supported financial aid, and more than $162.5 billion in aid from all sources, to offset postsecondary educational expenses (College Board, 2008a). In 2007–2008, two-thirds (66%) of all undergraduates received some type or amount of financial aid; for these recipients the average award was $9,100 (Wei et al., 2009).

The continued stratification of college opportunity even for academically qualified students despite the substantial annual investment of dollars into financial aid programs suggests that either: (1) existing financial aid resources are insufficient to meet students’ entire financial need; and/or (2) existing financial aid programs are not constructed or implemented as effectively as possible.

Various indicators support the former explanation. About half of all 2003–2004 undergraduates had some amount of unmet financial need, where unmet need is defined as total price of attendance less expected family contribution and all types of financial aid; the average amount of unmet financial need for undergraduates in 2003–2004 was $5,300 (Berkner & Wei, 2006). Undergraduates with the lowest family incomes are most likely to have some amount of unmet financial need, as the share of dependent undergraduates in 2003–2004 with unmet financial need ranged from 85% for those with family incomes below $20,000 to 12% for those with family incomes of $100,000 or more (Berkner & Wei, 2006). Even at public 2-year colleges, 87% of dependent undergraduates with family incomes below $20,000 averaged $4,500 in unmet financial need (Berkner & Wei, 2006).

Despite the apparent need for additional financial aid resources, however, federal and state governments are challenged to increase their investment in student aid programs. These challenges are typically exacerbated during economic downturns. The National Bureau of Economic Research (2008) determined that the U.S. economy began a recession in December 2007, defining a recession as “a significant decline in economic activity spread across the economy, lasting more than a few months, normally visible in production, employment, real income and other indicators” (p. 1). One consequence of a recession is a decline in government revenues. A December 2008 report from the Center on Budget and Policy Priorities projected that at least 41 states will have mid-year shortfalls in their FY2009 budgets (totaling $42 billion) and projected that nearly all states will have shortfalls in FY2010 (totaling $145 billion) and FY2011 (totaling $180 billion) (McNichol & Lay, 2008).

Most states are required to balance their budgets. As a result, when faced with revenue shortfalls states must utilize reserves, raise taxes, and/or reduce expenditures. Some (e.g., National Center for Public Policy and Higher Education, 2009) have urged states and higher education institutions to avoid responding to economic shortfalls through actions that increase the financial burden of college attendance on students and their families. Nonetheless, in response to shortfalls in their FY2009 budgets, at least 26 states have implemented or are considering reductions for higher education (McNichol & Lay, 2008). State budget shortfalls typically have a disproportionately negative effect on state appropriations for public higher education institutions, as higher education has historically served as a “balance wheel” in state budgets (Hovey, 1999). State appropriations for higher education are especially vulnerable in times of revenue shortfalls, as many legislators believe that colleges and universities have access to alternative sources of revenue, including tuition and fees, and may reduce personnel expenditures and realize other cost savings more easily than other state agencies (Lasher & Sullivan, 2004).

At the same time, federal and state governments are not only experiencing revenue shortfalls but also increasingly calling for accountability in the use of available resources. The federal government now requires that “effectiveness” be considered in decisions about the funding and management of federal programs. Reflecting the perspective that “federal programs should receive taxpayer dollars only when they prove they achieve results,” the U. S. Office of Management of Budget (2003) now rates federal programs as effective, moderately effective, adequate, or ineffective using the Program Assessment Rating Tool (PART). The PART evaluation produces an overall rating of effectiveness based on program performance in four areas: purpose and design (20% of total score), strategic planning (10% of total score), management (20% of total), and results and accountability (50% of total). Results and accountability must be demonstrated by measures of outcomes, outputs, and efficiency. Use of the PART ratings has been phased in, beginning with 130 federal programs for the 2003 budget and 234 programs for 2004, and will include all programs by 2008 (U.S. Office of Management and Budget, 2003). For the FY2009 budget, the Office of Management and Budget (2008, September) designated several student financial aid programs as “adequate” (e.g., Federal Family Education Loans, Federal Pell Grants, William D. Ford Direct Student Loans), meaning that these programs must “set more ambitious goals, achieve better results, improve accountability or strengthen its management practices” (U.S. Office of Management and Budget, 2009). The Federal Perkins Loan program was designated “ineffective,” meaning that the program has not demonstrated “results due to a lack of clarity regarding the program’s purpose or goals, poor management, or some other significant weakness” (U.S. Office of Management and Budget, 2009). Still other federal student aid programs (e.g., Byrd Honors Scholarships, Federal Work Study, Leveraging Educational Assistance Partnerships) were categorized as “results not demonstrated,” meaning that these programs do not have acceptable performance goals or data to assess performance (U.S. Office of Management and Budget, 2008).

Characteristics of Student Financial Aid

A first step toward developing a more complete understanding of the effects of financial aid on students’ college enrollment and how these effects may be mediated by context is to consider the complexities of available aid programs. Four relevant dimensions of student financial aid are: the source, the form of aid, goals, and eligibility criteria.

Source

Financial aid is available from many sources. In 2007–2008, about three-fourths of all student aid was awarded to undergraduates ($106.7 billion of $143.4 billion) rather than graduate or professional students; about two-thirds of this aid was awarded through federal programs (College Board, 2008a). Other providers of financial aid are colleges and universities (21% of all aid to undergraduates in 2007–2008), state governments (7%), and private organizations and employers (7%) (College Board, 2008a). In 2007–2008, 47% of all undergraduates received some type or amount of federal aid, 17.5% received some amount of state aid, and 21% received some amount of institutional aid (Wei et al., 2009).

With a few recent exceptions (e.g., Tebbs & Turner, 2006; Perna, 2008a), most research focuses on federal and state grants and federal loans with relatively little attention to understanding the effects on enrollment of scholarships from colleges and universities and other non-governmental sources (corporations, foundations, civic and community organizations and churches). Aid from these other sources may be substantial, as suggested by one electronic provider’s claim of having a database of “1.3 million scholarships worth over $3 billion” (Fastweb, 2009).

Form of Aid

Student financial aid comes in many different forms, including loans, grants, work-study, and tax credits and deductions (Hearn, 2001). The annual growth rate in loans, including non-federal loans, to undergraduates has exceeded the annual growth rate in grants to undergraduates since 1991–1992 (College Board, 2008a). As a result, loans represented 49% of all aid (including non-federal loans) received by undergraduates in 2007–2008, up from about 32% in 1991–1992 (College Board, 2008a). Colleges and universities are the largest source of aid in the form of grants, as 42% of all grant aid in 2007–2008 was from colleges and universities, 31% was from federal grants (i.e., federal Pell grants and other federal grant programs), 15% was from private and employer grants, and 12% from state grants (College Board, 2008a). About 6% of all aid to undergraduates in 2007–2008 was in the form of federal education tax credits (e.g., Hope and Lifetime Learning tax credits) and tax deductions, and 1% was in the form of federal work-study (College Board, 2008a). Looked at another way, in 2007–2008, 39% of all undergraduates received student aid in the form of loans, 52% in the form of grants, 7% in the form of work-study, and 2% in the form of veterans benefits (Wei et al., 2009). The average loan amount for those who received loans ($7,100) was higher than the average grant amount for those who received grants ($4,900) (Wei et al., 2009).

Goals of Aid Programs

Student financial aid programs are designed to achieve a range of goals (Hearn, 2001; Perna, Rowan-Kenyon, Bell, Li, & Thomas, 2008). Based on his examination of federally-sponsored financial aid programs, Hearn concluded that federal aid policies lack “philosophical coherence,” as reflected by the wide array of distinct goals, including promoting access for low-income students, improving college affordability for middle-income students, rewarding achievement, advancing economic development, and encouraging human capital investment. Perhaps reflecting the absence of “well-considered patterns of policy development” for federal student aid programs, Hearn also observed that taken together, federal student aid policies lack “programmatic clarity and distinctiveness.” In other words, based on his review of the literature, Hearn concluded that, “instead of an array of clearly discrete programmatic efforts addressing in distinctive fashion a set of overarching policy objectives, constituents for the programs… confront an array of overlapping efforts with rather vaguely differentiated objectives” (p. 270).

Similarly, based on a review of programs sponsored by the federal government and five states (California, Florida, Georgia, Maryland, and Pennsylvania), another study (Perna et al., 2008) identified distinctive state approaches to college-enrollment programs. Even for just these five states, the analyses showed variations in the extent to which college-enrollment programs include attention to addressing financial barriers and/or other barriers to college, recognize the role of schools in promoting students’ college enrollment, limit eligibility for aid based on financial need and/or academic achievement, and attempt to influence students’ decision to enroll in college versus their choice of institution to attend or some other outcome. Along the same lines, McDonough and colleagues (2007) concluded from their review of grant programs in 11 states that state approaches to financial aid vary in ways that reflect a state’s policy goals and objectives and that these variations likely have differential effects on college access and attainment.

Eligibility Criteria

Differences in program goals are manifest in part by differences in eligibility criteria, particularly the relative importance of measures of financial need and/or academic achievement. Over the past decade, the share of state financial aid awarded based on criteria other than financial need has increased substantially (College Board, 2008a). Between 1996–1997 and 2006–2007, the amount of non-need-based state grant aid awarded to undergraduates increased in constant dollars by 250%, while the amount of need-based state grant aid to undergraduates increased by only 59% (College Board, 2008a). Some states consider both financial need and academic achievement; 18% of all need-based state grant aid awarded in 2006–2007 also included academic criteria (College Board, 2008a). Although the majority of state grant aid is still need-based, and need-based state aid continues to increase annually, the faster rate of growth for merit- than need-based state aid raises concerns about the extent to which scarce public resources are being invested in ways that increase financial access for students from low-income families.

Summary of Current Knowledge Regarding the Effects of Financial Aid on College Enrollment

Others have provided outstanding reviews of research examining the effects of financial aid on college enrollment. Most notably, the Rethinking Student Financial Aid project sponsored by the College Board recently issued recommendations for reforming federal student aid along with a 220-page review entitled, “The Effectiveness of Student Aid Policies: What the Research Tells Us” (Baum, McPherson, & Steele, 2008). Several conclusions may be drawn from the Rethinking Student Aid literature reviews (Baum et al., 2008) and other relevant sources identified below.

First, the effects of student financial aid on college enrollment vary based on the type of aid. Based on their review of research examining the effectiveness of student financial aid, Baum and McPherson (2008) concluded that existing research devotes “strikingly uneven” attention to the effects on student outcomes of different types of aid programs (p. 5). Nonetheless, research consistently shows that grants promote college enrollment, especially for students from low-income students (Avery & Hoxby, 2004; Heller, 1997; Kane, 1999; Mundel, 2008). In contrast, while some studies (e.g., Avery & Hoxby, 2004) suggest that an offer of loans is positively related to enrollment net of other variables, Heller (2008) concluded that prior research generally shows loans to have minimal effect on college access and success. Although some research has examined the effects of work-study on student persistence in college (e.g., Pascarella & Terenzini, 2005; St. John, 2003), little is known about the effects of work-study on students’ college enrollment. In one of the few available studies, Avery and Hoxby (2004) found that, among high-ability high school seniors nationwide, the amount of work-study offered was positively related to the likelihood of enrolling after controlling for other student and institutional characteristics. The small number of available studies suggests that federal tax policies, including tax credits, deductions for higher education expenses and student loan interest, and tax-preferred savings programs, are unrelated to postsecondary access and choice (Long, 2004b; Reschovsky, 2008).

Second, research suggests that the effects of financial aid on college enrollment depend on a student’s family income and race/ethnicity (Avery & Hoxby, 2004; Heller, 1997; Kane, 1999; Mundel, 2008). Specifically, research consistently shows that changes in tuition and financial aid have a larger effect on college enrollment for students from lower-income families than higher-income families and for African Americans and Hispanics than for Whites (Avery & Hoxby, 2004; Heller, 1997; Kane, 1999; Long, 2004a). Some evidence suggests that the effects of financial aid on students’ choice of institution to attend also vary by race. Kim (2004) found that, net of other variables, receiving grants was associated with a higher probability of attending the first-choice institution for Whites, and receiving loans was positively related to attendance for Asians, while neither grants nor loans were related to the likelihood of attending the first choice institution for Blacks or Hispanics (Kim, 2004).

Nonetheless, although suggesting relationships between financial aid and college enrollment, existing research has several limitations. One limitation, as noted by others (e.g., Mundel, 2008; Heller, 2008), is that, because of methodological restrictions, existing research does not determine student outcomes in absence of the aid program.

A second limitation is that most research focuses on the relationship between a specific financial aid program and one college-related outcome, without considering how the broader context of student financial aid programs may mediate this relationship. This narrow focus is consistent with the goal of understanding whether a particular program “works.” However, research that does not consider the broader financial aid context may generate misleading conclusions. For example, in his review of research on the effects of loans on student outcomes, Heller (2008) recognizes the importance of considering the financial aid context when interpreting the results of available research. Specifically, Heller concludes that, although research generally finds that loans have a limited effect on college enrollment, this finding must be interpreted in light of the availability of grant aid:

This finding needs to be understood in the context of a financial aid landscape in which many students do not receive sufficient grant aid to pay for college. If grant aid were proportionally higher, then loans might provide more of a positive impact on college participation. But absent sufficient grant aid, simply piling on higher amounts of borrowing to students with large levels of unmet financial need may not be an effective vehicle for getting them to college (p. 49).

Similarly, Reschovsky (2008) concludes from his literature review that efforts to understand the effects of a specific federal tax policy on college outcomes are complicated by “the complex interactions among and between deductions, credits, subsidized savings plans, and direct higher education subsidy programs” (p. 76).

A third limitation is the absence of attention to the ways other aspects of context may mediate the relationship between student aid and college enrollment. Efforts to isolate the effects of a particular financial aid program on college enrollment should recognize the role of other policies aimed at promoting college enrollment. A few recent studies illustrate the ways multiple aspects of the state context come together to improve students’ college enrollment. For example, in an exploration of the forces that contributed to the substantial increase in college enrollment and other college-related outcomes in Indiana over the past two decades, Erisman and Del Rios (2008) concluded that Indiana’s success was driven by a commitment to, and progress toward attaining, four goals: (1) improving high school students’ academic readiness for college by making the college preparatory curriculum the “default” for all high school students; (2) increasing college affordability especially for low-income students; (3) increasing the diversification of the state higher education system by expanding the role of community colleges; and (4) promoting college success by rewarding institutions for student outcomes and developing policies to facilitate transfers across institutions in the state. Efforts to improve college affordability included attention to the cost of public higher education institutions, availability of need-based state grants (i.e., Frank O’Bannon Grants), dissemination of information about the availability of financial aid (e.g., Learn More Indiana website), early commitment of financial aid and academic and social supports through the Twenty-first Century Scholars Program, and establishment of the Part-Time Grant program to promote enrollment of non-traditional students (Erisman & Del Rios, 2008).

Other analyses, using different methods, also suggest the need to consider how the state policy environment and other state characteristics may mediate the effects of any particular aid program on enrollment. Using interrupted time-series analyses of data from the Integrated Postsecondary Education Data System (IPEDS), Orsuwan and Heck (2009) found that state merit aid was inversely related to the percentage of freshmen who attended out-of-state colleges and universities. But, the magnitude of the relationship depended on other characteristics of the state, including whether the state also adopted a prepaid tuition program and tuition rates at the state’s public flagship university relative to tuition at public flagship universities nationwide (Orsuwan & Heck, 2009).

A fourth limitation of available research is the failure to consider how the effects of aid may be mediated by characteristics of individual students and their families, such as what students and their families know about student aid. Existing research tends to examine the effects of actual amounts of aid on student enrollment, with little consideration of students’ knowledge or perceptions of the aid. Despite the availability of financial aid, many reports document the absence of accurate knowledge of financial aid among high school students and their families (e.g., Chan & Cochrane, 2008; Grodsky & Jones, 2004; Horn, Chen & Chapman, 2003; Ikenberry & Hartle, 1998). National surveys consistently show that most adults, parents, and students are uninformed or poorly informed about college prices and financial aid (Grodsky & Jones, 2004; Horn et al., 2003; Ikenberry & Hartle, 1998). Levels of awareness and understanding of college prices and financial aid are particularly low among Latino students and parents (Immerwahr, 2003; Tomás Rivera Policy Institute, 2004).

A related limitation is that available research typically considers only the direct effects of a financial aid award on enrollment, without examining how knowledge or perceptions of financial aid may indirectly promote college enrollment. As others (e.g., Fitzgerald, 2006; Paulsen & St. John, 2002) have argued, financial aid may promote college access not only directly by reducing or eliminating financial barriers at the time of college entry but also indirectly by encouraging students to engage in other college-related behaviors, including other stages of the college-enrollment process, because they know aid will be available.

Based on their review and synthesis of prior research, Hossler and Gallagher (1987) concluded that the three stages of the college process are predisposition, search, and choice. In the first stage, students become predisposed toward or interested in attending college as they develop educational and occupational aspirations (Hossler & Gallagher, 1987; Terenzini, Cabrera, & Bernal, 2001). In the second stage, students search for information about colleges (Hossler & Gallagher, 1987; Terenzini et al., 2001). While still less frequently researched than the other two stages, researchers who have examined this stage typically operationalize “search” in terms of the sources of college-related information that students and parents use (e.g., Hossler, Schmit, & Vesper, 1999) and/or the number of colleges that students consider or to which they apply (e.g., Hossler et al., 1999; Hurtado, Inkelas, Briggs, & Rhee, 1997). In the third stage, students decide to enroll in a particular college or university. Little is known about the timing of these three stages for non-traditional enrollment. But, for “traditional” college enrollment (i.e., enrollment into college immediately after graduating from high school), predisposition typically occurs between the 7th and 10th grades, search during the 10th–2th grades, and choice during the 11th and 12th grades (Hossler et al., 1999; Terenzini et al., 2001).

Finally, existing research does not devote sufficient attention to the ways that student and family demographic and cultural characteristics may mediate the effects of aid on college enrollment. As an example, research demonstrates that responsiveness to aid varies across groups based on race/ethnicity and family income (Avery & Hoxby, 2004; Heller, 1997; Kane, 1999; Mundel, 2008), but does not explain why these differences occur. Potential explanations may include cultural differences in perceptions of aid, as well as differences across groups in access to resources that promote knowledge and understanding of financial aid.

In short, while providing a baseline of knowledge, existing research on the effects of student financial aid on college enrollment provides insufficient attention to the context in which the programs are implemented or the ways various dimensions of context may mediate the effects of aid on enrollment. Moreover, existing research does not consider the ways financial aid may indirectly promote college enrollment. Addressing the limitations of prior research requires the use of a conceptual model that explicitly considers the role of context. The following section offers a potential conceptual model.

Conceptual Framework for Understanding the Role of Context

In earlier work (Perna, 2006a), I drew on a review and synthesis of prior research to propose a multi-layered conceptual model of college enrollment. This prior work identified financial aid as one force that contributes to college enrollment but did not focus specifically on the relationship between aid and enrollment. Drawing on this earlier work as well as St. John’s (2003) balanced access model, this chapter argues that the relationship between financial aid and college enrollment is mediated by various dimensions of context. Like “the student choice construct” (Paulsen & St. John, 2002; St. John, Asker, & Hu, 2001), the conceptual model assumes that college enrollment decisions reflect an individual’s “situated context” and that pathways to college enrollment differ in ways that reflect the diversity in individual circumstances, as well as the ways that individual circumstances serve to define and constrain students’ college opportunities.

Drawing on the economic theory of human capital, the model assumes that students make decisions about college enrollment based on an assessment of the lifetime benefits and costs of enrollment relative to alternative choices (Perna, 2006a). Consistent with this economic approach, students are assumed to make choices that maximize their utility. Reflecting sociological theoretical perspectives, the model also assumes that students’ college enrollment decisions occur within multiple layers of context. The four nested contextual layers proposed in the model are: students and their families; K-12 schools; higher education institutions; and the broader societal, economic, and policy context (Perna, 2006a).

The conceptual model is designed in part to illustrate the multiple ways policymakers may intervene to promote college enrollment (Perna, 2006a). By specifying linkages between policies and college-related outcomes, St. John’s (2003) balanced access model provides additional insights into the ways public policy interventions such as student financial aid shape college enrollment behavior. In his framework for assessing the influence of policy on educational opportunity, St. John identifies several key steps in the educational attainment process: K-12 attainment and achievement, postsecondary transitions and access, undergraduate and graduate student outcomes, and individual development and educational attainment. St. John’s framework posits that K-12 policies pertaining to schooling and school reform (e.g., standards and testing) shape K-12 attainment and achievement, policy interventions (e.g., financial aid policy, postsecondary information, and affirmative action) shape postsecondary transitions and access, and college and university policies (e.g., financial and academic strategies) shape undergraduate and graduate student outcomes. The relationship between financial aid and college enrollment may be understood from both a macro- and micro-economic perspective. Figure 1 shows how the two perspectives may be considered together.

Fig. 1

Conceptual model of student college enrollment with policy linkages (Source: Perna et al., 2008)

Financial Aid as a Lever for Increasing Student Demand for Higher Education (Macro-Level)

At the macro-level (i.e., a public sector finance perspective), student financial aid is conceptualized as a “lever” that may be used to intervene in the “market” for higher education with the goal of promoting college-related outcomes (Paulsen, 2001a). More specifically, student financial aid policies are designed to increase students’ demand for higher education, and thus their college enrollment, by increasing the supply of resources that students have available to invest in higher education (Paulsen, 2001a).

Economists argue that government intervention in the higher education market, including intervention in the form of student financial aid, is warranted for at least four reasons (Kane, 1999; Paulsen, 2001a). First, individual participants do not capture all of the benefits that higher education produces, as the benefits “spillover” to non-participants. Because individuals make enrollment decisions based on an assessment of only the individual benefits without considering the benefits that accrue to society, the level of investment in higher education that would occur without market intervention is less than optimal (Paulsen, 2001a; Steuerle, 2001). By providing student financial aid to reduce the costs of enrollment to individual participants, governments attempt to address this tendency to under-invest in higher education and maximize the societal benefits of higher education (Hansen & Weisbrod, 1969). The societal benefits of higher education are well-documented (Baum & Ma, 2007; Bowen, 1997; Institute for Higher Education Policy, 1998), and include increased economic productivity, tax revenues, and community and civic engagement, as well as reduced crime rates and dependence on public welfare.

Government-sponsored financial aid programs, particularly student loan programs, are also designed to correct for a second market failure: the inability of participants to use their post-higher education level of human capital as collateral against which to borrow funds needed to pay educational prices (Kane, 1999; Paulsen, 2001a). In the absence of market intervention, private capital markets fail to provide funds students need to pay college prices, again resulting in an under-investment in higher education. Guaranteed student loan programs address this limitation by providing incentives for lenders to make funds available for students to borrow and allocating funds to students in the form of grants.

A third justification for government intervention is that individuals not only lack perfect information about college opportunities but also have differential access to such information (Kane, 1999). As with some other goods (e.g., restaurant meals, shoes), buyers of higher education are unable to obtain complete information about the “product” until they “experience” it (Winston, 1999). Potential first-generation college students, a disproportionate percentage of whom are from low-income families and are Black or Hispanic (National Center for Education Statistics, 2001), are likely to be particularly disadvantaged by this characteristic of higher education markets since they cannot rely on their parents for relevant information. Governments may address this market failure by disseminating information about the availability of financial aid programs.

Government intervention in the higher education market is also warranted when the public supports a reduction in inequities across groups (Paulsen, 2001a). Paulsen argues that public polices that increase higher education enrollment are a more efficient approach to equalizing incomes than other government interventions including direct transfers to low-income individuals or those from underrepresented racial/ethnic groups. By building human capital, student aid and other policies that promote college enrollment enable individuals to earn higher incomes throughout their lifetimes, thereby eliminating the need for the public sector to provide annual income subsidies to these individuals.

Government-sponsored student aid programs are typically designed to directly influence students’ demand for college enrollment by increasing the supply of resources they have available to pay college prices (Paulsen, 2001a). One review revealed that, although the federal and state governments have distinct approaches to college-enrollment policy, 89% of all college-enrollment policies sponsored by the federal government and five selected states (California, Florida, Georgia, Maryland, and Pennsylvania) include only a financial component and are designed to impact students’ college enrollment directly (Perna et al., 2008). For example, with Pell Grants the federal government aims to directly increase low-income students’ college enrollment by providing need-based grants that reduce the price of attending college. Only a minority of programs include components to address other barriers to college enrollment (e.g., academic preparation and information), and relatively few programs explicitly recognize the role of high schools in implementing financial aid programs.

Financial Aid as a Mechanism for Increasing a Student’s Supply of Resources (Micro-Level)

At the micro-economic level, human capital theory assumes that students decide to enroll in college by comparing the lifetime benefits and costs for all alternatives and then selecting the alternative (e.g., college, work) that maximizes the student’s utility (Becker, 1993; Paulsen, 2001b). This perspective predicts that the provision of financial aid directly to students will increase the supply of resources a student and his/her family have to pay for college and consequently increase the likelihood a student will conclude the benefits of attending college outweigh the costs.

Limitations of Human Capital Theory

Nonetheless, as others have noted (e.g., Long, 2009; Paulsen, 2001b; Perna, 2006a), although conceptually appealing human capital theory alone is insufficient for completely understanding how financial aid influences students’ college enrollment decisions. First, human capital models do not explain several inconsistencies in students’ responses to financial aid (Long, 2009). For example, human capital models do not explain why students “react differently to various forms of financial aid and tuition changes, even if the economic value of each is the same” (Heller, 1997, p. 632). Human capital theory also does not explain why college enrollment is influenced by non-pecuniary aspects of financial aid, including whether the aid is labeled “grant” or “scholarship” or whether the grant aid is frontloaded (Avery & Hoxby, 2004).

A second limitation is that, although acknowledging differences in students’ preferences for different types of aid, human capital theory does not explore how or why preferences are formed (Manski, 2009). A rational human capital investment model assumes that, even when the expected benefits and costs are the same, two individuals may make different college choices because of differences in their preferences, tolerance for risk, and uncertainty (DesJardins & Toutkoushian, 2005). But, human capital theory does not explore why preferences, tastes, and expectations differ across groups (DesJardins & Toutkoushian, 2005).

Therefore, although useful for conceptualizing the effects of costs and benefits on students’ college enrollment behaviors, traditional human capital approaches alone are insufficient for understanding inconsistencies in college-enrollment decision-making or how students develop perceptions or understandings of financial aid. Drawing on other frameworks provides additional insights.

Other Theories

Both behavioral economics and psychological approaches to decision making assume that, because of informational and computational constraints on information processing capacities, individuals adopt such strategies as satisficing or bounded rationality (Hogarth, 1987; Long, 2009). Thus, because of cognitive limitations, individuals make decisions that are bounded by time and resource constraints.

Drawing on a sociological perspective, McDonough (1997) uses Bourdieu’s notion of habitus to illustrate the ways that bounded rationality influences college-related decision making. This perspective assumes that college-related decisions reflect an individual’s habitus, or the internalized system of thoughts, beliefs, and perceptions that is acquired from the immediate environment. Rather than consider all possible alternatives, habitus defines and limits the alternatives that are considered, how different alternatives are perceived and valued, and the choices that are made (McDonough, 1997; Paulsen & St. John, 2002). These alternatives are defined and delimited by the social, organizational, and cultural contexts in which individuals are embedded, including class-based values and the organizational habitus of high school attended (McDonough, 1997). According to McDonough, the notion of organizational habitus considers both the availability of resources and structures within a high school to promote college enrollment, as well as the underlying norms and expectations for college enrollment at a school, norms and expectations that reflect class-based values about college-going.

Other sociological theories, such as cultural and social capital, may also be useful for understanding differences in students’ perceptions and use of financial aid. Consideration of cultural and social capital provides a more complete assessment of the resources that students have available from their social contexts to inform college-related decisions (McDonough, 1997). Cultural capital refers to the system of attributes, such as language skills, cultural knowledge, and mannerisms, that is derived, in part, from one’s parents and that defines an individual’s class status (Bourdieu, 1986; Bourdieu & Passeron, 1977). Members of the dominant class possess the most economically and symbolically valued kinds of cultural capital (Bourdieu & Passeron, 1977; McDonough, 1997). McDonough uses the notion of “entitlement” to further understand the role of cultural capital in college-going decisions. For example, students attending elite private high schools feel “entitled” to highly-selective colleges while students attending urban high schools in working-class neighborhoods feel “entitled” to community colleges (McDonough, 1997).

Social capital focuses on social networks and the ways social networks and connections are sustained (Morrow, 1999). In his comprehensive assessment of the origins and uses of social capital, Portes (1998) noted that social capital is acquired through an individual’s relationships with other individuals, particularly through membership in social networks and other social structures. Coleman’s (1988) approach stresses the role of social capital in communicating the norms, trust, authority, and social controls that an individual must understand and adopt in order to succeed. Whereas Coleman’s perspective suggests that parents play a primary role in promoting the status attainment of their children, Bourdieu’s approach describes the restrictions that structural barriers to institutional resources, in the form of differential access across racial/ethnic, gender, and other groups, impose (Dika & Singh, 2002). According to Bourdieu (1986), the amount of social capital to which an individual may gain access through social networks and relationships depends on the size of the networks as well as the amounts of economic, cultural, and social capital that individuals in the network possess.

The Importance of Context in Understanding the Effects of Aid on Enrollment

Among other conclusions, these theories and the research that tests them illustrate the ways that various aspects of “context” influence postsecondary access and choice. Students make postsecondary access and choice decisions largely determined by the norms and values embedded within a student’s family and high school, as well as a broader economic, social, and policy context (Perna, 2006a). Greater attention to context may help identify the structures and processes that define and delimit students’ perceptions and use of financial aid in college-enrollment decisions. These understandings, in turn, may suggest ways to improve the construction, implementation, and/or marketing of student financial aid programs to more effectively promote enrollment of all students.

Other researchers assert the benefits of considering context when examining students’ college-related behaviors (e.g., Orsuwan & Heck, 2009; Tierney & Colyar, 2006). McDonough, Calderone, and Purdy (2007) argue that attention to context may be especially useful for understanding inconsistencies in students’ financial aid-related decisions:

Reinterpreting affordability as a localized, highly contextualized, deliberative process sheds potential light on why low-income students fail to claim state and federal aid even when they qualify, as well as why large-scale attempts to disseminate financial aid information, streamline the FAFSA process, and improve practitioner knowledge related to financial aid have resulted in minimal improvements to the college-financing perceptions of those most in need (pp. 5–6).

With this conceptual model as a guide, the following section suggests the ways that the student and family, school, higher education, and broader economic, social, and policy contexts may mediate the relationship between student financial aid and college enrollment.

Applying the Conceptual Model to Understanding the Ways the Effects of Aid on College Enrollment are Mediated by Context

The Student and Family Context

An emerging body of research suggests that perceptions of aid vary by race/ethnicity and family income (Cunningham & Santiago, 2008; Hahn & Price, 2008; Heller, 2008; Linsenmeier, Rosen, & Rouse, 2006; Perna, 2008b). Differences across groups in willingness to borrow are typically inferred based on students’ actual use of loans. For example, using regression discontinuity and difference-in-difference analyses, Linsenmeier et al. (2006) found that an institution’s decision to eliminate loans from financial aid packages for low-income students increased the likelihood of enrollment for low-income, minority students but was unrelated to the likelihood of enrollment for low-income students overall. The authors speculated that replacing loans with grants had a greater effect on enrollment for minority than non-minority applicants because minority students are more averse to debt than non-minority students.

Research also suggests that willingness to borrow varies by family income, although the nature of the relationship is ambiguous. Some descriptive (e.g., Baum & O’Malley, 2003; Perna, 2008b) and multivariate (e.g., Callender & Jackson, 2005; Linsenmeier et al., 2006) analyses find that students from low-income families are less willing than other students to borrow, while multivariate analyses (e.g., Christou & Haliassos, 2006; Eckel, Johnson, Montmarquette, & Rojas, 2007) show that students from low-income families are more likely than other students to borrow.

Several aspects of the student and family context may help explain differences across racial/ethnic and family income groups in perceptions of aid, as well as differences in the effects of aid on enrollment. One potentially relevant dimension of the student and family context pertains to social values and cultural norms about loans (Cunningham & Santiago, 2008; Heller, 2008). For example, using data from the 2004 National Postsecondary Student Aid Survey and BPS:04/06, Cunningham and Santiago (2008) found that minority students, especially Asian and Hispanic students, were more likely than other students to have at least $2,000 in unmet financial need but low loan amounts. Using focus groups to probe this finding, Cunningham and Santiago concluded that Asian and Hispanic students may be less willing to borrow, at least in part, because of cultural norms against borrowing.

A second potentially relevant dimension for understanding differences in the use and perceptions of aid may pertain to a family’s prior experience with financial aid, including the experiences of parents and older children (Long, 2009; McDonough, 1997; Perna, 2008b). Consistent with cultural and social capital theories, data from one exploratory study show that students’ perceptions of loans are closely related to their parents’ views (Perna, 2008b).

A related dimension of the student and family context pertains to the relative roles of students and their families in paying college prices (Long, 2009; Manski, 2009; Perna, 2006b). An exploratory study suggests that some students do not apply for available scholarships (Perna, 2008a) or learn the specifics of student financial aid programs (Perna & Steele, in press) when they believe that their parents will fund the costs of their postsecondary education.

Over the past decade, the share of college costs paid by parents has declined, while the share paid by students has increased (Hearn, 2001; Stringer, Cunningham, O’Brien, & Merisotis, 1998). Although many factors (e.g., increased consumer debt, inadequate savings, slow personal growth) may contribute, one source of the decline in the share of costs covered by parents may be parents’ reduced willingness to pay (Stringer et al., 1998).

Some research suggests variations across racial/ethnic and socioeconomic groups in terms of the expected distribution of college costs between parents and students. In her case study analyses of college-related decisions of students attending four high schools, McDonough (1997) found that students in high-SES families typically defer college financing decisions to their parents but that students in low-SES families generally assume responsibility for identifying ways to pay college prices and more explicitly consider college prices and financial aid in college application and enrollment decisions. Using national data and multinomial logit analyses, Steelman and Powell (1993) found that, even after controlling for parents’ education, parents’ marital status, number of children, and family income, African American, Hispanic, and Asian parents were more likely than White parents to perceive college costs to be the responsibility of parents rather than students. Based on their analyses, Ellwood and Kane (2000) concluded that college enrollment rates may be positively related to family income at least in part because parental willingness to pay college prices increases with family income.

Although most families want to promote their child’s educational attainment, the ability of low-income and minority families to do so is often constrained by economic, social, and psychological barriers (Perna, 2004). Barriers to parental involvement may be associated with mother’s employment, parents’ education, family composition, child care responsibilities, language and culture, and discrimination, as well as a lack of prior experience with college-related processes (Ceja, 2001; Kerbow & Bernhardt, 1993; López, Scribner, & Mahitivanichcha, 2001; Tierney & Auerbach, 2005). In the face of these barriers, some families rely on their children to obtain college-related information (Ceja, 2001; Tomás Rivera Policy Institute, 2004; Tornatzky, Cutler, & Lee, 2002). Children are then responsible not only for acquiring and understanding information about college prices and financial aid, but also for educating their families about this information (Ceja, 2001).

School Context

A central aspect of school context that may mediate the effects of aid on enrollment is the availability of financial aid and other college-related counseling from school staff. Qualitative research illustrates that school counselors play an important role in structuring students’ aspirations, plans, and readiness for college (McDonough, 1997, 2004, 2005a, 2005b). Using case study analyses of four high schools in California, McDonough (1997) identified variations across schools not only in the availability of college counseling but also in counselors’ assumptions about students’ other sources of college-related knowledge, “appropriate” college destinations for students attending the school, and ways to best disseminate college-related information to students.

Research also suggests the role of school counselors in shaping students’ perceptions of financial aid (McDonough & Calderone, 2006; Perna, Rowan-Kenyon, Thomas, et al., 2008). Based on her review of available research, McDonough (2004) concluded that students and parents rely on school counselors as a source of information about financial aid (when counselors are available), and that students who talk with “available and trained” school counselors are more knowledgeable about college prices than other students.

Support from high school counselors may be especially important when parents do not have the knowledge, information, prior direct experience, or other resources required to adequately guide their children through college-related processes (Furstenberg, Cook, Eccles, Elder, & Sameroff, 1999; Kerbow & Bernhardt, 1993; McDonough, 1997; Tierney & Auerbach, 2005). Compared to White students and parents, African American, Hispanic, and low-income students and parents appear to be more dependent on high school personnel for information about college (Cabrera & La Nasa, 2001; Ceja, 2001; Freeman, 1997; Horn et al., 2003; Terenzini et al., 2001; Tomás Rivera Policy Institute, 2004; Tornatzky et al., 2002; U. S. GAO, 1990). However, the high schools these students typically attend are generally not equipped to provide support for college admissions-related activities to all students (McDonough, 1997; Rosenbaum, 2001). The availability of college counseling varies across schools not only in terms of actual student-to-counselor ratios, but also in terms of time devoted to college counseling, relationships between school counselors and college admissions staff, and types of counseling services offered, as well as the availability of other school structures to support college counseling, such as a school’s mission and norms regarding college-going and degree of college-preparatory curricular focus (McDonough, 1997). The ability of school counselors to engage in college counseling is further limited by the need for counselors to engage in other activities including crisis intervention counseling, developmental counseling, scheduling, test administration, and discipline (Ballard & Murgatroyd, 1999; McDonough, 2005a, 2005b; NACAC, 2006; Venezia & Kirst, 2005). One exploratory study suggests that these resource constraints reduce the availability of counselors for one-on-one meetings, shift the focus of counseling to the needs of the school’s “typical” or most “needy” students, and require students and their families to initiate contact with school counselors (Perna, Rowan-Kenyon, Thomas, et al., 2008).

School counselors appear to be especially unavailable and ill-equipped to talk with students about financial aid (McDonough, 2004, 2005a, 2005b). Financial aid training is typically not part of a school counselor’s formal education, as college counseling has traditionally been viewed as inconsistent with a counselor’s focus on students’ mental health (McDonough, 2005a, 2005b). Although counselors typically provide information about financial aid as part of their duties, only one in four counselors responding to a NACAC (2006) survey felt “very prepared” to guide students’ borrowing decisions. Counselors are particularly unsure about advising students about the amount to borrow, the optimal type of loan, or the consequences of failing to repay loans (NACAC, 2006).

Other research raises questions about the quality or comprehensiveness of financial aid information school counselors provide. Few counselors appear to systematically disseminate information to students about financial aid (McDonough & Calderone, 2006; Perna, 2008a, 2008b). Drawing on interviews and focus groups with 63 college counselors at urban high schools in southern California, McDonough and Calderone (2006) found that, with a few exceptions, most counselors provided no more than minimal information or assistance to students regarding college financing, most encouraged students to attend low-price institutions (e.g., community colleges) based on their assumptions about students’ ability to pay, and some counselors advised students based on the assumption that African American and Latino parents are reluctant to use loans to finance college prices. Similarly, an exploratory study of the forces that contribute to college-going among students attending 15 high schools revealed that high school counselors and teachers provide a range of messages to students about loans, with some discouraging and others encouraging their use (Perna, 2008b).

Higher Education Context

Variations in both the prevalence of institutional grant aid and the criteria for awarding institutional grant aid suggest that colleges and universities may also mediate the effects of financial aid on college enrollment. The share of undergraduates receiving financial aid varies by institutional type. In 2007–2008, 48% of undergraduates attending public 2-year institutions received some type or amount of financial aid, compared with 70% of undergraduates at public 4-year colleges and universities, 87% of undergraduates at private, not-for-profit non-doctorate-granting institutions, and 98% at private for-profit, 2-year or 4-year institutions (Wei et al., 2009). Average amounts of financial aid for recipients also vary by institutional type, ranging from $3,400 at public 2-year institutions, to $10,100 at public 4-year doctorate-granting institutions, to $19,000 at private 4-year doctorate granting institutions (Wei et al., 2009). On average, private 4-year institutions award a higher share of aid to students based on financial need than public 4-year institutions. In 2006–2007, 70% of all institutional grants at private 4-year institutions were awarded based on financial need, compared with only 44% at public 4-year institutions (College Board, 2008a). Public 4-year colleges and universities also tend to devote a larger share of institutional aid to athletic awards than private 4-year colleges and universities (18% versus 6%, on average, in 2006–2007, College Board, 2008a).

Private 4-year colleges and universities charging higher tuition also tend to allocate a higher percentage of institutional grant aid based on financial need than those charging lower tuition, regardless of Carnegie classification (College Board, 2008a). For example, high-tuition private doctoral-granting institutions allocated 80% of their institutional aid based on financial need, while their low-tuition counterparts devoted just 61% (College Board, 2008a). Institutional grants tended to cover a higher percentage of published tuition and fees at lower-priced than higher-priced private institutions (e.g., 34% versus 18% at higher- and lower-priced private baccalaureate colleges) (College Board, 2008a).

Even among highly selective private colleges and universities financial aid and pricing polices vary. In an examination of financial aid records of students attending 28 highly-selective private colleges and universities, Hill, Winston, and Boyd (2005) found that, on average, net prices increased with family income; net-price as a share of median family income ranged from 49% for students in the lowest family income quintile to 21% for students in the 95th percentile of family income. But, the ratio of net price to family income varied substantially, ranging, for the lowest-income students, from 5 to 74% (Hill et al., 2005). At seven of the 28 institutions, net price represented a larger percentage of income for low-income students than for high-income students (Hill et al., 2005).

Social, Economic, and Policy Context

Many aspects of the broader social, economic, and policy context may mediate students’ perceptions of aid and the effects of financial aid on enrollment. Two particularly relevant aspects of this broader context are state variations in student financial aid and changes in the characteristics of available financial aid. Variations in student financial aid across states suggest the importance of considering the characteristics of aid programs, as well as other state characteristics, when examining the effects of aid on student outcomes (Perna & Titus, 2004; Orsuwan & Heck, 2009). Changes in the characteristics of financial aid programs over time suggest limitations on the generalizability of findings from prior research (Heller, 2008; Hossler Hossler, Ziskin, Sooyeon, Osman, & Gross, 2008).

Variations in State Financial Aid. Research shows that state grant aid is positively related to college enrollment, regardless of whether the aid is awarded based on need or merit. For example, the likelihood of enrolling in any type of postsecondary education within 2 years of graduating from high school and the likelihood of attending an in-state private 4-year or in-state public 4-year college or university increase with the availability of state need-based financial aid (Kane, 1999; Perna & Titus, 2004). Using fixed-effect regression analyses of state-level data from 1992 to 2000, St. John, Musoba, and Chung (2004) found that, after controlling for state demographic and other characteristics, both need-based and non-need-based grants were positive predictors of college enrollment among high school graduates; a $1,000 increase in need-based grants was associated with an 11.5% point increase in college enrollment rates, while a $1,000 increase in non-need-based grant aid was associated with an 8.9% point increase.

Research also suggests that the effects of state merit-aid programs on enrollment vary in ways that reflect the characteristics of the aid program. For example, studies show increases in college enrollment associated with the state merit-aid program in Georgia (Cornwell, Mustard, & Sridhar, 2004; Dynarski, 2000, 2002, 2004) but not in New Mexico (Binder, Ganderton, & Hutchens, 2002). Dynarski (2004) also found that the Georgia HOPE Scholarship has a greater effect on enrollment at public 4-year institutions than at other types of institutions. In a comparison of seven state merit-aid programs, Dynarski (2004) found the largest enrollment effects in Georgia, Kentucky, Louisiana, and Mississippi and the smallest in South Carolina (where the relationship was not statistically significant). The effects of merit-aid programs on racial/ethnic gaps in enrollment also varied across states (Dynarski, 2004). Merit-aid programs in three states had larger effects on enrollment for Blacks and Hispanics than for Whites, thereby reducing racial/ethnic gaps in enrollment. In contrast, the Georgia HOPE Scholarship program had larger effects for Whites than for Blacks and Hispanics, suggesting that the program contributes to the racial/ethnic stratification of enrollment in the state (Dynarski, 2004).

Other research suggests that the effects of state aid on college enrollment may be mediated by other state characteristics. For instance, using interrupted time-series analysis of data from IPEDS, Orsuwan and Heck (2009) concluded that measures of a state’s economic and political characteristics mediated the positive effects of a state merit-aid program on the out-migration of college freshmen. Out-migration was positively related to per-capita income and negatively related to the percentage of state education expenditures allocated to higher education. The reduction in out-migration associated with having a state merit-aid program was even larger in states that also had prepaid tuition programs and relatively higher rates of tuition at the state’s public flagship institutions than in other states (Orsuwan & Heck, 2009).

Changes in Financial Aid Over Time. The availability and purchasing power of financial aid also likely influence the relationship between aid and enrollment. These types of changes over time stem from several sources. Some occur with the periodic reauthorization of the Higher Education Act of 1965, the authorizing legislation for federal student aid programs. For example, the 2008 reauthorization of the Higher Education Act (aka the Higher Education Opportunity Act) broadened eligibility for the federal Academic Competitiveness and Smart Grants to include students enrolled part-time and in certificate programs (Field, 2008). The 1992 reauthorization included changes that increased the amounts students could borrow and relaxed some aspects of the needs-analysis criteria. These changes led to substantial increases in borrowing and thus limit the generalizability of findings from research examining the use of loans prior to 1992 (Heller, 2008).

Changes in the purchasing power of student financial aid programs have occurred in response to changes in government resources. For example, although the Higher Education Act of 1965 authorizes the Federal Pell Grant program, the federal government must annually appropriate resources for the awards. As result, maximum and average Pell grant awards have fluctuated over the past three decades. In the past decade alone, the maximum Pell grant (in 2007 dollars) rose and fell from $3,504 in 1997–1998, to $4,626 in 2002–2003, to $4,146 in 2006–2007, and $4,310 in 2007–2008 (College Board, 2008a). As another example, in response to funding concerns, Georgia legislators tightened academic eligibility requirements for the Georgia HOPE Scholarship in 2004, effective for students entering college on or after May 1, 2007 (Fisher, 2007). Because of this change, the number of 2007 high school seniors eligible for HOPE declined by about 18,000 (Fisher, 2007).

Changes in economic conditions may restrict the availability of student aid. For example, one implication of the current credit crisis is that 130 fewer lenders offered federal student loans and 31 fewer lenders offered private loans at the start of 2008–2009 than at the start of 2007–2008 (Thomson, 2008). A second implication of the current economic recession and subsequent budget constraints is that some states are seeking to reduce expenditures by restricting the availability of state student aid programs. For example, in December 2008, New Jersey state legislators voted to reduce expenditures for two merit-based state aid programs by making eligibility requirements more restrictive (Giordano, 2008). The changes require high school students to graduate in the top 15% rather than the top 20% of their class in order to receive the NJ STARS scholarship, a program established in 2004 to cover tuition and fees at any New Jersey community college. With the revisions, the scholarship will no longer cover tuition and fees for remedial coursework. The changes also require STARS recipients to graduate from a community college with at least a 3.25 grade point average rather than a 3.0 grade point average to receive NJ STARS II, a scholarship to attend a 4-year public college or university, restrict eligibility for NJ STARS II to students with family incomes below $250,000, and limit the scholarship award from full-tuition to $6,000 per year for students with grade point averages between 3.25 and 3.49 and $7,000 for students with grade point averages above 3.49 (Giordano, 2008).

The purchasing power of financial aid also changes over time due to changes in the price of attendance. Over the past two decades, the sticker price of attending all types of colleges and universities has increased dramatically (College Board, 2008b). Between 1998–1999 and 2008–2009, average tuition and fees increased by 27% at private 4-year colleges and universities, 50% at public 4-year colleges and universities, and 15% at public 2-year colleges and universities after controlling for inflation (College Board, 2008b). The extent to which increases in financial aid have kept pace with increases in tuition has also varied over time. For example, the share of average tuition and fees at public 4-year colleges and universities that is covered by the maximum Pell grant was 50% in 1987–1988, declined to 35% in 1996–1997, rose again in the early 2000s, and fell again to 32% in 2007–2008 (College Board, 2008a).

Questions for Future Research

The proposed conceptual model suggests the ways various aspects of context mediate the relationship between financial aid and students’ college-related behaviors. Considering the limitations of prior research described earlier in this chapter, the proposed conceptual model suggests at least five broad questions for future research. Addressing these questions with a conceptual model that explicitly considers the role of context will likely generate insights into how to maximize the effectiveness of available financial aid resources.

How Do Different Types of Aid Separately and Together Influence College Enrollment?

One area for future research is to better understand how different types of aid separately and together influence college enrollment. As described in the review of existing research above, one limitation of most research is the tendency to focus on isolating the effects of only grants or loans, with relatively little attention to the effects on college enrollment of other forms of aid (Baum & McPherson, 2008) or how different forms of aid interact to influence students’ college-related behaviors (Orsuwan & Heck, 2009). The “purest” forms of financial aid are grants and scholarships, i.e., funds that may be used to pay postsecondary educational expenses and do not have to be repaid. About half (50%) of all undergraduates enrolled in 2003–2004 received financial aid in the form of grants; the average amount of grant aid received was $4,000 (Berkner & Wei, 2006).

Other forms of aid, including loans, work-study, and tax credits, have “strings” attached. Unlike grants, loans need to be repaid. Moreover, even within “loans,” program terms vary. In 2007–2008, 34% of all loans were federal subsidized Stafford Loans, 31% federal unsubsidized Stafford Loans, 13% federal PLUS loans (for parents of dependent undergraduates), and 23% non-federal loans (College Board, 2008a). The College Board notes that federal loans are a form of aid but non-federal loans are not, as non-federal loans include no interest subsidy and “generally have less favorable terms than federal loans” (p. 9). Unlike non-federal loans, federal Subsidized and Unsubsidized Stafford Loans include a fixed-rate of interest and allow for deferment of repayment until 6 months after graduation. Unlike Unsubsidized Stafford Loans, Subsidized Stafford Loans are awarded based on a student’s financial need and include an interest “subsidy,” i.e., the federal government, not the student, pays the interest while a student is enrolled at least half-time.

Federal work-study also has “strings,” as students must work in designated positions in order to realize the award. A form of campus-based federal aid, eligible colleges and universities may award federal work-study to financially needy students. About 1% of all aid to undergraduates in 2007–2008 was in the form of federal work-study (College Board, 2008a). Other forms of financial aid have increased substantially over the past decade. For example, between 1997–1998 and 2007–2008, total federal grants increased in constant dollars by 79% and total federal loans increased by 70%. In contrast, over this period the total amount of federal work-study dollars awarded to undergraduates has remained virtually unchanged after controlling for inflation (College Board, 2008a).

Federal education tax credits and tuition deductions have characteristics that differentiate them from other forms of aid. Whereas other federal student aid programs (e.g., Pell grants, Stafford Loans, Federal Work-Study) are authorized and periodically reauthorized under the Higher Education Act of 1965, the federal Hope and Lifetime Learning tax credits were established as part of the federal tax code (i.e., the Taxpayer Relief Act of 1997). In 2007–2008, 6% of all aid to undergraduates was in the form of federal education tax credits and tuition deductions (College Board, 2008a). Observers point to several reasons for the absence of an enrollment effect for federal tax policies, including the delay in receiving the award (i.e., in the year after tuition expenses were paid), the failure of tax credits to cover costs of attendance other than tuition and fees (e.g., room, board, books, other educational supplies), the limitation of the credits based on tax liability (i.e., tax credits are not refundable), and the complexity of the programs and tax forms (Long, 2004b; Reschovsky, 2008).

States also offer a range of programs that are designed to increase the resources that students have to pay for college. These programs include not only state-sponsored financial aid programs, but also prepaid tuition and college savings plans, tuition reciprocity agreements with other states, and loan forgiveness programs, among others (Orsuwan & Heck, 2009). Prepaid tuition and college savings plans are tax-advantaged vehicles that encourage families to save for their children’s postsecondary educational expenses, but by definition require families to have sufficient disposable income from which to save.

Given the complexity of financial aid, it is not surprising that most research tends to consider only broad categories of particular forms of aid (e.g., “grants” or “loans”). Nonetheless, in order to understand how to structure financial aid so as to maximize its impact on enrollment, future research should give greater attention to the complexities of aid, including variations in the effects of aid based on the sources, types, goals, and eligibility criteria. Future research should also consider the ways the effects of any particular financial aid program are mediated by the context for aid, including the availability and characteristics of other financial aid programs (Orsuwan & Heck, 2009). Because of the tendency to examine particular financial aid programs in isolation, little is known about how the effects of a particular financial aid program vary based on the availability of other types of financial aid (Heller, 2008; Reschovsky, 2008).

How Do Perceptions of Aid Influence Students’ College-Related Decisions?

Future research should also consider how the effects of financial aid are mediated by student and family characteristics, particularly their knowledge and perceptions of aid. Rational human capital models do not assume that individuals have perfect and complete information about all alternatives, only that individuals use available information to make “reasoned” choices (DesJardins & Toutkoushian, 2005, p. 218). Despite this acknowledgement, and as described in the review of prior research above, human capital models typically consider actual amounts of grants and loans rather than students’ knowledge or perceptions of financial aid (Long, 2009; Manski, 2009). Considering only actual amounts of financial aid is especially problematic since, with the exception of some state merit-aid programs (e.g., Georgia HOPE), students and their families only learn the amounts and types of financial aid they will receive after they have applied for admission and financial aid (Kane, 1999; Perna & Steele, in press).

Some evidence suggests that perceptions of loans vary across groups and that these perceptions are related to college-related behaviors. For example, although the growth of student loans suggests that the use of loans to finance postsecondary educational expenses is commonplace (College Board, 2008a), the emphasis of the U.S. financial aid system on loans may limit college opportunity for individuals who are unwilling or unable to incur this type of debt (Perna, 2008b). Although some research suggests that students’ perceptions of loans vary across groups (e.g., Christie & Munro, 2003; Christou & Haliassos, 2006; ECMC Group Foundation, 2003; Linsenmeier et al., 2006), only a few studies have explored the forces that contribute to these differences (e.g., Christie & Munro, 2003; Perna, 2008b).

Some research suggests that willingness to borrow is positively related to college enrollment (Callender & Jackson, 2005; Ekstrom, 1991). Using data from the High School and Beyond longitudinal survey of 1980 high school sophomores and seniors, Ekstrom found that students who reported they would be willing to borrow to pay $1,500 in college prices that could not be covered by family or other sources of financial aid were more likely than other students to enroll in college within 4 years of graduating from high school, enroll in a 4-year than a 2-year institution, and enroll full-time than part-time even after controlling for background characteristics, educational aspirations, academic achievement, encouragement from significant others, and knowledge of college costs and financial aid. Using data from surveys of 2,000 prospective higher education students in Britain, Callendar and Jackson found that the likelihood of applying to a university increased with students’ tolerance for debt even after controlling for educational achievement, social class, ethnicity, age, and mother’s educational attainment. Moreover, debt aversion had a greater negative relationship to the probability of applying for admission for students from low- than high-income families (Callender & Jackson, 2005).

Although these studies suggest the importance of students’ perceptions of aid, additional research is required to understand the ways these perceptions influence students’ college-related behaviors. Future research should further explore students’ perceptions of financial aid, the forces that influence these perceptions, and the implications of these perceptions for student outcomes. Future research should also examine how perceptions of financial aid are informed by various aspects of the context in which students are embedded, including their family background, support from school staff, the characteristics of available aid programs, and other forces including the media.

What Types of Information About Financial Aid, at What Points in Time, from What Sources, and for What Groups of Students Promote College Enrollment?

Little is known from prior research about how financial aid indirectly promotes enrollment, including how knowledge of financial aid may raise students’ college-aspirations or encourage students to become academically prepared for college. Therefore, a third area for future research is to consider how financial aid—and particularly information about financial aid-influences these college-related behaviors and how the relationship between financial aid information and college-related behaviors is mediated by various aspects of context.

Various policies and practices are aimed at increasing students’ and families’ college-related knowledge. Among the most recent is the public service advertising campaign, KnowHow2Go campaign that was launched in January 2007 by the American Council on Education, Lumina Foundation for Education, and the Ad Council. This campaign is designed to disseminate information about the steps required to enroll in college, including how to “put your hands on some cash.”

Nonetheless, as indicated in the discussion of the limitations of existing research above, despite the availability of information dissemination efforts reports consistently demonstrate that most students and parents are poorly informed (Perna, 2004). One benefit of improving knowledge of financial aid may be to reduce the number of students who enroll in college and are eligible for financial aid but do not apply. Using descriptive analyses of data from the National Postsecondary Student Aid Survey, King (2004) found that about 1.7 million low- and moderate-income undergraduates enrolled in colleges and universities nationwide in 1999–2000 had not completed the FAFSA; about half of these students were likely eligible for Pell grants. A substantial number of other undergraduates completed the FAFSA but submitted it after April 1, the preferred deadline for most state and institutional grant aid (King, 2004). Additional descriptive analyses of data from the Beginning Postsecondary Student survey suggest that many eligible students do not submit a FAFSA because of incomplete or inaccurate understandings of their eligibility for aid and submission deadlines (King, 2004).

Despite the belief that “information” about financial aid promotes college enrollment, little is known about the necessary content, timing, and/or modes of delivering messages about financial aid (Mundel & Coles, 2004; Perna, 2004). Therefore, future research should identify what types of knowledge “matter” (i.e., promote various college-enrollment behaviors), at what points in time, from which sources, for which groups of students. If “information” is to serve as an effective policy lever for increasing college access and choice, then we need to know more about whether information can promote changes in student behavior and what kind of information different groups of students need to have at different points in the pathway to college.

Types of Information. While descriptive data consistently illustrate gaps in knowledge about college prices and financial aid (Perna, 2004), future research should identify the types of knowledge that are important. For example, do students need to have accurate and complete knowledge of college prices and financial aid, or do they more simply need to be confident that, when it comes time to enroll, they’ll be able to secure the necessary resources? One exploratory study Perna & Steele (in press) suggests that knowledge of financial aid per se may be less important in encouraging college-enrollment behaviors than the presence of family, school, and other supports and structures that engender confidence in students that they will be able to pay.

Timing of Information. Future research should also examine when students and their families need to have information about financial aid. Several recent reports call for “early” information about financial aid, asserting that students and research now tend to learn about financial aid only during the later years of high school, i.e., too late to influence other college-related behaviors (Advisory Committee on Student Financial Assistance, 2008; Perna, 2004).

Nonetheless, little is known about how the timing of information, including an early promise of aid, shapes college-related behaviors. Based on a review of available research, Schwartz (2008) concluded that little is known about the effects of early commitments of aid, due to limitations of available research. Schwartz identifies four types of financial aid programs that make an “early commitment” to students about the future availability of aid: (1) state merit-aid grant programs that award aid based on high levels of academic achievement; (2) I Have a Dream, Indiana’s Twenty-first Century Scholars Program, and other programs in which students self-select to participate; (3) stakeholder grants such as the Child Trust Fund in the United Kingdom; and (4) other programs, including individual development accounts (IDAs). Research examining the effects of early commitment programs is limited by methodological issues (e.g., self-selection of students and families into these programs) and programmatic issues (e.g., limited participation in most programs of students from low-income families, limited dollar amounts of most awards).

Knowledge of financial aid is believed to be related to students’ college-enrollment decisions (Perna, 2004), as research shows that students’ and parents’ understanding of college prices and financial aid is positively related to college expectations (Flint, 1993; Horn et al., 2003), application (Cabrera & LaNasa, 2000), enrollment (Plank & Jordan, 2001), and choice (Ekstrom, 1991), as well as such college financing strategies as students’ willingness to borrow, students’ use of financial aid, parental saving for college (Ekstrom, 1991; Flint, 1997), and student application for financial aid (U. S. Government Accounting Office, 1990).

Available research does not, however, establish the direction of causality between knowledge of financial aid and students’ college-related behaviors (Perna, 2004). In other words, existing research does not establish whether having knowledge of financial aid causes students to engage in college-related behaviors, or whether those who engage in these behaviors acquire information and knowledge. Available research also does not establish the relationship between knowledge of aid and other college-related behaviors, including aspirations for college prior to the 11th grade, curricular choices, college search processes, or students’ choice among institutions (Mundel, 2008).

Sources of Information. Future research should also consider how, and with what consequence, the effects of information about financial aid are mediated by various aspects of context—particularly the sources providing financial aid information. The proposed conceptual framework suggests that potential sources of information may include other family members (e.g., older siblings and extended family members), high school staff, peers attending the same high school, pre-college outreach programs (e.g., Upward Bound), colleges and universities, and federal and state sponsors of student financial aid. Available research includes limited attention to only some of these potential sources.

School staff. Perhaps in response to what is known from available research, some have begun to advocate for improving the quantity and quality of college counseling by high school counselors. For example, one of the ten recommendations made by the College Board’s Commission on Access, Admissions and Success in Higher Education (2008) is to increase the availability of college counseling during middle and high school by reducing student-to-counselor ratios, improving preparation and training for college and financial aid counseling, and maximizing the use of other resources to provide counseling, including local colleges and universities and community-based non-profit organizations.

As others have noted (e.g., McDonough, 2004; Mundel, 2008), although existing research suggests the importance of counselors as a source of financial aid information, more research is required to understand the ways they influence students’ and families’ knowledge of financial aid. As McDonough (2004) notes, although research suggests the benefits to college enrollment of timely and educated discussions between students and school counselors, little is known about school counselors’ knowledge of aid.

Additional research is also required to understand how and with what consequences other school staff may be providing financial aid counseling to students. The ability of teachers to serve as a source of information about college costs and financial aid is often limited because teachers are often focused on other priorities, including reducing high school dropout rates and teen pregnancies (Immerwahr, 2003), and/or lack knowledge about college-related requirements beyond their own direct experience (Venezia, Kirst, & Antonio, 2003). Teachers also often have low educational expectations for African Americans and Hispanics (Freeman, 1997; Immerwahr, 2003). Nonetheless, in the absence of sufficient financial aid counseling from school counselors, at least some students may be relying on information from teachers (Perna, Rowan-Kenyon, Thomas, et al., 2008).

Future research should also examine how the provision of financial aid information varies based on the complexity and other characteristics of student aid programs (Mundel, 2008; Perna, 2004). One exploratory analysis suggests that, although state merit-aid programs in Florida and Georgia may have negative consequences for equity, the relative simplicity of eligibility may enable counselors and teachers to more confidently communicate with students about the availability of and criteria for receiving financial aid (Perna, Rowan-Kenyon, Thomas, et al., 2008). In three other study states (i.e., California, Maryland, and Pennsylvania), school staff did not provide extensive financial aid assistance to students at least in part because of the complexity of need-based aid application processes (Perna et al., 2008).

Higher education institutions. More research is needed to understand how colleges and universities affect students’ understanding and use of financial aid. Variations across institutions in the presence of Pell-grant recipients suggest that the effects of aid on college enrollment depend on institutional characteristics. Using regression analyses of IPEDS data to explore this variation, Steinberg, Piraino, and Haveman (2009) found that the representation of Pell-grant recipients among undergraduates was inversely related to the institution’s median SAT score at both public and private institutions. The magnitude of the relationship was substantially greater at public than private institutions. At private institutions, the percentage of Pell-grant recipients was also inversely related to the total cost of attendance.

More research is needed to understand the effects of higher education institutions’ efforts to “market” financial aid on students’ perceptions and use of financial aid (Mundel, 2008; Perna, 2004). Merely by their presence and geographic proximity to students’ homes, higher education institutions may passively convey some information about college to students and their families (Leppel, 1994; McDonough, Antonio, & Trent, 1997). Geographic proximity may be a proxy for greater availability of information about an institution and greater knowledge of the institution among a student’s family, school, and community (Leppel, 1994; McDonough et al., 1997).

Higher education institutions may actively convey information to students and their parents through targeted marketing and recruiting efforts (Chapman, 1981). Most such actions are likely only reactive, however, occurring in response to some type of initiative by the student, including taking college admissions examinations and applying for admission (Cabrera & LaNasa, 2001). Because students with low socioeconomic status are less likely than other students to apply for admission (Cabrera & La Nasa, 2001), the “reactive” ways in which higher education institutions convey information may contribute to differences across groups in knowledge of various aspects of college including financial aid.

Higher education institutions actively convey information about college prices and financial aid to students and their families through the financial aid notification process (Kane, 1999). But colleges and universities typically do not inform students and their families of their eligibility for financial aid until they have applied and been accepted for admission, applied for student financial aid, and been deemed eligible for some amount and type of aid. In other words, students and their families typically only learn about the amounts and types of financial aid that they will receive after they have made substantial investments in the college-going process (Heller, 2006). Along the same lines, some research suggests that financial aid offices of colleges and universities may influence the loan-related experiences of students who actually enroll (Wroblewski, 2007). Although mandatory entrance and exit counseling may improve borrowers’ understanding of some aspects of loans (Wroblewski, 2007), such counseling comes too late to influence high school students’ perceptions of loans (as only students who successfully apply for and receive admission and financial aid and decide to borrow receive entrance counseling).

Future research should examine the effects on perceptions and use of financial aid of recent efforts by some colleges and universities to replace loans with grants for students from low-income families. Most of these programs have been established by Ivy League institutions (e.g., Princeton; University of Pennsylvania), public flagship universities (e.g., University of North Carolina Chapel Hill; University of Maryland College Park), and elite private liberal arts colleges (e.g., Amherst, Davidson, and Williams Colleges), institutions with relatively small numbers of low-income students (McPherson & Shapiro, 2006). Research should not only consider the effects of these institutional aid programs on enrollment and choice for low-income students, but also examine how these effects vary based on student characteristics and other institutional characteristics. Among the potentially important institutional characteristics may be the ways that an institution communicates the availability of this aid program to prospective students.

More research is needed to understand how students’ perceptions of institutional financial aid, including the lack of knowledge of need-based aid, may lead students to self-select out of particular colleges. One exploratory study suggests that at least some high school students and parents have a particularly sophisticated understanding of institutional scholarships, with some planning to attend the college that makes the best institutional aid offer (Perna, 2008a). But the findings also suggest that other parents believe that, because colleges and universities target aid to students’ with the lowest family incomes and highest academic achievement, relatively few resources are available for “average” students (Perna, 2008a).

Finally, additional research is required to understand how colleges and universities may work with high schools to best inform students about the availability of student financial aid (Perna, Rowan-Kenyon, Thomas, et al., 2008). One exploratory study suggests that both schools and higher education institutions benefit from the common practice of higher education staff conducting an annual financial aid night at local high schools; schools benefit from the provision of financial aid information by local experts and colleges and universities benefit from direct access to potential applicants and their families. By working to identify other collaborative opportunities, schools and higher education institutions may not only advance their own goals but also maximize the availability of college and financial aid counseling in the context of scarce resources.

Federal, state, and local providers. Several recent reports argue the need to simplify, and increase the transparency of, financial aid application processes (e.g., Advisory Committee on Student Financial Assistance, 2005, 2009; Baum & McPherson, 2008; Chan & Cochrane, 2008; Commission on Access, Admissions, and Success in Higher Education, 2008; Secretary of Education’s Commission on the Future of Higher Education, 2006). But little research has considered the other ways federal, state, and local providers of financial aid may promote or deter knowledge of financial aid. Future research should examine the utility of other mechanisms for disseminating information about financial aid. One potential type of intervention that warrants further consideration is the effort to match available scholarship dollars to eligible students. The Georgia Career Information System (GCIS) offers a potential model for providing students with one source for information about all available scholarship dollars and matching students to scholarship dollars (Perna, 2008a). The GCIS (2007), housed at Georgia State University, is designed to “provide current and accurate career information to schools and agencies throughout Georgia in order to help young people and adults make informed occupational and educational choices.” Among other types of information, the GCIS “contains information on more than 3,567 federal, state, and independent aid programs representing about 5 billion dollars” (GCIC, 2007).

Technology. Little is known about the effects on students’ and families’ knowledge and use of financial aid of Internet sites or other mechanisms for disseminating information about financial aid, including financial aid workshops, required college and career counseling sessions, or one-on-one counseling sessions with college alumni and others (Porter et al., 2006). Lumina Foundation for Education (2009) urges the Obama administration to work toward the following goal: “By the 8th grade, provide all children and families with information needed to plan for college success.” But Lumina goes on to note that KnowHow2Go and other public awareness campaigns alone are insufficient to ensure that students have the knowledge required to enroll in college. Lumina is now focusing on “develop[ing] a ground campaign to link youth to local resources of caring adults and organizations to help students take the steps needed, including developing stronger state and local college access networks.” This action suggests a disillusionment with the potential benefits of relying on technology alone as an information source. Given the prevalence of financial aid-related Internet sites, future research should consider the ways technology, with and without other types of support, may provide financial aid information to different groups of students and families (Mundel, 2008).

How Can Financial Aid Promote College-Related Outcomes Other than College Enrollment, Particularly Students’ Academic Preparation for College?

A fourth area for future research is to consider how another aspect of context—namely the characteristics of aid—may directly and indirectly promote college enrollment. More specifically, given that college enrollment and success are limited not only by insufficient financial resources but also by inadequate academic preparation (Perna, 2006a), future research should examine the ways to construct aid programs so as to both encourage academic achievement and eliminate financial barriers (McDonough, Calderone, & Purdy, 2007).

Little is known about the effects of federal efforts to use grant aid to improve college readiness, largely because these efforts are still relatively new. Established by then-president Bush in FY2006, the Academic Competitiveness Grant (ACG) and Smart Grant programs award supplemental Pell grant funds to financially needy students who complete a rigorous curricular program in high school, or major in mathematics or science in college. The number of recipients of ACG and Smart grants was higher in 2007–2008, the second year of the program, than in 2006–2007 (465,000 versus 370,000, Field, 2009). Nonetheless, the number of participants and total spending on the program were still lower than expected in 2007–2008. In 2006–2007, about $430 million was awarded to students through the two programs, substantially less than the $790 million Congress had appropriated (Field, 2009). The future of these programs is unclear as President Obama’s proposed FY10 budget would allow these programs to expire after 2010 (Lederman, 2009).

Little is also known about the effects of programs with multiple eligibility criteria on student outcomes, even though state aid programs are increasingly including both academic merit and financial need criteria (McDonough et al., 2007). McDonough and colleagues underscore the policy-relevance of research that informs decisions about the criteria for awarding financial aid by stating:

As a society, we lack a comprehensive student-aid policy that attempts to provide a threshold of financial aid that truly enables poor students to afford college while providing the appropriate proportion of aid that encourages and rewards meritorious performance (p. 24).

One goal of many merit-based state grant programs is to improve students’ educational attainment by increasing their academic readiness (Doyle, 2008). Nonetheless, based on a review Doyle concluded that available research does not establish whether merit-aid programs achieve this result. Similarly, Dynarski (2004) concluded from her research review that state merit aid programs are associated with improved academic performance in high school and college, but that other forces may also explain these trends.

Some research suggests a complex relationship between state merit-aid programs and academic preparation. For example, one exploratory study suggests that Georgia’s HOPE Scholarship program may motivate students to achieve the academic requirements of the state grant award (Perna & Steele, in press). By clearly communicating academic eligibility criteria, these programs may increase parents’ attention to students’ academic preparation (Perna & Steele, in press). This study also suggests potential negative consequences of state merit-aid programs for academic readiness, such as increasing pressure on teachers to inflate grades and encouraging students to take less-rigorous courses (i.e., where they are more likely to meet the grade point average requirement for the merit aid).

More research is needed on other potential unintended consequences of state merit-aid programs for academic preparation. Based on his review of research, Mundel (2008) speculates that, with the apparent simplicity and transparency of state merit-aid programs, some students may conclude that the academic eligibility criteria are impossible to achieve and consequently lower their educational aspirations and forgo other college-preparatory activities (Mundel, 2008). Others may conclude that financial aid is available only based on academic merit, without appropriately recognizing the availability of need-based aid (Mundel, 2008).

One commonly observed disadvantage of state merit-aid programs is that these programs disproportionately benefit students from upper-income families and others who would have attended college even without the aid (Heller, 2004; Heller & Marin, 2002). Given these well-documented challenges, the College Board’s Commission on Access, Admissions and Success in Higher Education (2008) declares that “merit aid, particularly when financed publicly by regressive taxes or lotteries, has to clear a very high bar before it can justify itself as appropriate merit aid” (p. 29).

Nonetheless, recent growth in state merit aid (College Board, 2008a) and the establishment of the federal ACG and Smart Grants suggest policymakers’ interest in using student aid to motivate students to become adequately academically prepared to enroll and succeed in college. Therefore, future research should further explore the ways financial aid eligibility requirements may be constructed to promote academic preparation and eliminate financial barriers while also minimizing any unintended negative consequences (McDonough et al., 2006).

Future research should also consider how other types of approaches to financial aid may also improve students’ academic performance and readiness for college. One particularly promising approach may be to provide an early commitment of student financial aid. Based on his review, Harnisch (2009) notes that state-based early-commitment programs have several characteristics including targeting of middle-school students and their families; a clear and simple contract that specifies the requirements for receiving financial aid; minimum high school grade point averages and college preparatory curricular requirements; and guaranteed financial aid for college. While some research examines the effects of Indiana’s Twenty-First Century Scholars on students’ college-related outcomes (e.g., St. John, Musoba, Simmons, & Chung, 2002; St. John, Gross, Musoba, & Chung, 2005; St. John, Fisher, Lee, Daun-Barnett, Williams, 2008), little is known about the effects of other state-based early-commitment programs including Oklahoma’s Promise, Wisconsin Covenant, and Washington College Bound Scholarship. Future research should capitalize on the variations across these programs to develop a better understanding of the implications of early-commitment programs for students’ academic preparation and other college-related behaviors. These four state-based early-commitment programs vary in terms of the grade level students must make the commitment (e.g., as early as sixth grade in Indiana and as late as 10th grade in Oklahoma); maximum income (e.g., free and reduced lunch status in Indiana and no income requirement in Wisconsin); and minimum high school grade point average (ranging from 2.0 in Indiana and Washington to 2.85 in Wisconsin) (Harnisch, 2009).

How Does Financial Aid Affect College Enrollment for Adult Students, and How is the Relationship Between Aid and Enrollment for Adult Students Mediated by Various Aspects of Context?

A final area for future research is to examine the effects of financial aid on college enrollment for an under-researched population, adult students, and how aspects of context mediate this relationship. Virtually all of the research and the conceptual model presented in this chapter focus on understanding the effects of financial aid on college enrollment for traditional-age students. This focus likely reflects, at least in part, the absence of research on the role of financial aid in college-going decisions for adult students, as well as the general absence of aid programs specifically targeted toward adult students (e.g., Lapovsky, 2008). In his review, Mundel (2008) concludes that little is known about the effects of grants on college-related behaviors of students who delay college enrollment after finishing high school.

Policymakers have begun to recognize the importance of improving college-related outcomes for adult students. As an example, recommendations from the U. S. Department of Education Margaret Spellings Commission on the Future of Higher Education (2006) include ensuring that individuals have the opportunity to participate in higher education over the course of their lifetimes. Nonetheless, based on her review, Lapovsky (2008) concluded that the existing financial aid system was not developed to promote college access among adult students. She speculates that some aspects of the financial aid system may favor adult students, as suggested by the high percentage of Pell Grant awards received by financially independent students. Other aspects, however, do not, such as the emphasis of state and institutional grant aid programs on full-time enrollment; a spring deadline for applying for many state and institutional aid programs; work disincentives in the needs-analysis formula; and the diversity of personal and enrollment characteristics among adult students (Lapovsky, 2008). Additional research is required to understand the ways these and other dimensions of context influence adult students’ use and perceptions of financial aid and their consequences for enrollment-related outcomes. Additional research should also consider the availability, use, and consequences of employer-provided financial aid for the enrollment of adult workers.

Recommended Strategies for Future Research

Further drawing on the limitations of prior research and the proposed conceptual model, this section offers four recommended strategies to guide future research: (1) develop research designs that will identify not just what “works” but how to improve financial aid programs; (2) recognize the contributions of multiple methodological approaches; (3) capitalize on differences among financial aid programs; and (4) recognize the contributions of multiple theoretical perspectives.

Develop Research Designs that Identify How to Improve Programs

In this era of accountability, demonstrating whether a particular financial aid program “works” is critical to a program’s future funding. In its discussion of the federal PART ratings, the U.S. Office of Management and Budget (OMB, 2003) clearly articulates the value it places on such research by stating, “No program, however worthy its goal and high-minded its name, is entitled to continue perpetually unless it can demonstrate it is actually effective in solving problems” (p. 47). OMB goes on to underscore this point by stating: “What works is what matters, and achievement should determine which programs survive and which do not” (p. 53).

Certainly program effectiveness should be considered when making decisions about the allocation of scare public resources. Nonetheless, there are at least three cautions associated with relying on measures of program effectiveness to determine whether a program is a good investment. One caution pertains to the interpretation of the findings from such research. In an article in the New England Journal of Medicine, Henry Aaron (2008) discusses the question, “What, exactly, is wasteful health care spending?” (p. 1). Like health care, investments in education programs likely have a continuum of benefits with the magnitude varying across individuals. Even when an investment in health care—or education is labeled as “wasteful” or “ineffective,” some individuals are likely benefiting. Given these benefits, whether we continue to offer these programs becomes an ethical decision (Aaron, 2008).

A second caution is that policymakers and practitioners need to know more than simply whether a program “works.” Clearly more information is also needed about how financial aid programs work and how they may be adjusted to maximize their effects on college-related outcomes (Mundel, 2008). For example, although most research focuses on whether grants are related to college enrollment, policymakers and practitioners would also benefit from knowing how grants directly and indirectly contribute to college enrollment, as well as how grant programs may be improved so as to increase the magnitude of the enrollment benefits and/or the number of individuals who benefit (Mundel, 2008).

Finally, a consideration of the limitations of prior research in light of the conceptual model demonstrates that understanding whether a financial aid program “works” is complicated by the complexities of the contexts in which these programs operate. Other researchers have recognized these complexities (Orsuwan & Heck, 2009). For example, Stout (2008) identifies the “confounding factors” that challenge efforts to isolate the effects of the Academic Competitiveness and SMART grant programs on student outcomes, including other federal, state, and local initiatives; the complexity of the programs; and changes in student eligibility and economic conditions. Considering the role of context suggests the challenges associated with determining whether a particular program causes a particular change in student outcomes, or what student outcomes would have occurred in the absence of the program.

Recognize the Contributions of Multiple Methodological Approaches

Although existing research examining the effects of financial aid on student outcomes has important analytic and methodological limitations (Heller, 2008), no single design or study is without flaws. When selecting a research design or methodology, researchers must acknowledge that all designs have limitations (Baum & McPherson, 2008).

Given the limitations inherent in any design, as well as the complexities of the questions that should be addressed in future research, multiple methodological approaches are necessary. Taken together, findings from multiple methodological approaches will produce a more comprehensive understanding of these complex questions. In particular, researchers should consider the insights that may be gained from randomized control trials, multi-level modeling, qualitative methods, and longitudinal designs.

Randomized Control Trials. According to the federal government, the “gold standard” for research is randomized control trials (RCTs), the only design that can establish whether a particular program causes a given outcome (U.S. Office of Management and Budget, February 2008). By randomly assigning students to “treatment” and “control” groups, randomized control trials, unlike other methods, are assumed to control for all other explanations for any differences in outcomes between the two groups (besides the treatment). As others (e.g., Heller, 2008; Long, 2004c) have noted, the heavy reliance in existing research on descriptive and quasi-experimental designs necessarily limits what is known about whether financial aid causes improvements in college enrollment.

Ethical and political challenges restrict the use of randomized control trials (Baum & McPherson, 2008). OMB (2008, February) recognizes these challenges by noting that “randomized controlled trials are generally the highest quality, unbiased evaluation to demonstrate actual impact, but [should be used] only when it is appropriate and feasible to conduct such studies” (slide 40). OMB goes on to note that “a variety of quasi-experimental methods (e.g., comparison group studies) and non-experimental methods may help shed light on how or why a program is effective;” “evaluations must be appropriate for the type of program” (U. S. Office of Management and Budget, 2008, February, slide 40).

Therefore, given the potential insights gained from establishing causal relationships, researchers should consider the ways RCTs may be utilized for examining the relationship between financial aid and college enrollment. When politically and ethically feasible, RCTs may offer important insights.

Multi-level Modeling. A second fruitful approach to future research examining the effects of financial aid on college enrollment is to use multi-level modeling (e.g., hierarchical linear modeling). By explicitly considering multiple levels, multi-level modeling may be particularly useful for understanding the effects on enrollment of multiple dimensions of context.

Some existing studies suggest the insights that may be generated through multi-level modeling. For example, using multilevel modeling, Perna and Titus (2005) explore the ways the characteristics of the high school attended influence the college enrollment decisions of high school graduates. Focusing specifically on the role of parental involvement as a form of social capital, Perna and Titus operationalize aspects of the school context in terms of the extent to which the school encourages parental involvement, the volume of resources that may be accessed via social networks at the school, and the homogeneity of these social networks.

Their analyses show that, regardless of an individual student’s social, economic, cultural, and human capital, the likelihood of enrolling in a 2-year or 4-year college after graduating from high school is related to the volume of resources that may be accessed through social networks at the school attended. The volume of resources is measured by such variables as the average levels of parental involvement, family income, parental education, and parental educational expectations at the school the child attends.

Multi-level modeling may also shed light on how the state context influences college enrollment. Perna and Titus (2004) use multilevel modeling to examine the effects of various types of state public policies on the type of college or university high school graduates attend after taking into account student-level predictors of enrollment. State-level variables included measures of state appropriations to higher education, tuition, availability of need-based and non-need-based student financial aid, K-12 education, and the availability of higher education in the state.

With relatively large sample sizes, high response rates, multiple data sources, national samples, and the clustering of students within high schools, the longitudinal datasets sponsored by the National Center for Education Statistics (NCES) are an important source of data for modeling effects at the student, school, and state levels (Perna & Titus, 2004, 2005). NCES is currently beginning the fourth in a series of longitudinal studies that provide data on students’ transition from high school to postsecondary education. The National Longitudinal Study (NLS) of 1972 high school seniors followed students periodically through 1986. The High School and Beyond (HS&B) Study followed 1980 high school seniors through 1986 and 1980 high school sophomores through 1992. The National Educational Longitudinal Study (NELS) contains data for a cohort of students in the 8th grade (1988) when most of the students were high school sophomores (1990), high school seniors (1992), 2 years after their scheduled high school graduation (1994), and 8 years after (2000). The sample was freshened in 1990 and 1992 to ensure representative cohorts of 1990 tenth graders and 1992 twelfth-graders, respectively. The Educational Longitudinal Study 2002 (ELS) tracks the experiences of 2002 tenth graders through high school and into postsecondary education and the workforce, with data collections in 2002, 2004, and 2006. The High School Longitudinal Study (HSLS:09) will follow 2009 ninth graders forward. The HSLS will include data from about 20,000 high school students, their parents, and school staff (i.e., teachers, guidance counselors, and administrators).

Qualitative methods. Existing knowledge of the effects of student financial aid on students’ college-related outcomes is largely informed by the results of quantitative research methodologies. A small but growing body of research, however, demonstrates the insights that may also be generated from qualitative research methodologies.

As others (e.g., St. John, 2006) note, qualitative methods may be especially useful for understanding students’ perceptions of financial aid and how these perceptions influence their college-related decisions. For instance, using semi-structured interviews with 49 students in the United Kingdom, Christie and Munro’s (2003) qualitative study offers useful insights into how students view debt and how financial and cultural resources influence their views.

In an examination of the college pathways of five students attending one urban high school in Los Angeles, Tierney and Colyar (2006) (along with their chapter authors) demonstrate the contribution of qualitative methods to understanding the ways students’ educational choices are influenced by their contexts. The authors use intensive qualitative data collection methods, including weekly meetings with each student over the course of 1 year, and “cultural biography” as a framework for identifying the ways cultural contexts influence high school students’ college pathways. Colyar (2006) argues that cultural biography is an understanding of students as situated within a particular culture and social context. Together, the portraits of these five students illustrate the ways college-going processes are informed by a student’s “cultural system,” including their family (e.g., encouragement to be first-generation college student), school (e.g., scarcity of resources), community (e.g., poverty), and other aspects of their context (Tierney, 2006).

Longitudinal research designs. Longitudinal designs are especially important for understanding how information and knowledge about financial aid influence not only students’ actual college enrollment but also other college-related behaviors including aspirations for college, academic preparation for college, search for and application to colleges, and choice among colleges (Long, 2009; Perna, 2004).

Longitudinal designs are especially important for understanding when students need to receive various types of financial aid information. Most agree that providing students and their families with information about college prices and financial aid early in the educational pipeline may encourage students to engage in college preparatory behaviors, such as becoming academically prepared for college (Perna, 2004; Baum & McPherson, 2008). Nonetheless, as explained in this chapter’s review of prior research, existing research does not indicate whether having information causes students to engage in college-related behaviors, or whether students who would have engaged in these behaviors anyway acquire information (Long, 2009; Perna, 2004). Data that longitudinally track changes in both knowledge and behavior would help to address this knowledge gap.

Longitudinal designs are also important for addressing other questions regarding the effects of financial aid on students’ college-related behaviors. For example, longitudinal designs are needed to better understand the extent to which students retain financial aid (especially merit-based aid) after enrolling and go on to successfully complete their degree programs (Perna & Steele, in press). As another example, Orsuwan and Heck (2009) demonstrate the ways that a longitudinal design (i.e., interrupted time-series analysis) may be used to examine the effects of adopting state merit-aid and prepaid tuition policies on the enrollment of freshmen in out-of-state colleges and universities.

Capitalize on Variations in Student Financial Aid

While complicating efforts to understand the effects of financial aid on college enrollment, variations in existing financial aid programs also offer an opportunity for researchers. In particular, future research might consider exploiting variations in state approaches to financial aid in order to better understand how various program characteristics influence college enrollment and choice.

Some research (e.g., Dynarski, 2004; Orsuwan & Heck, 2009; Perna & Steele, in press) suggests how differences in state approaches to financial aid may be used to better understand how different approaches influence students’ perceptions about and use of financial aid. State grant programs have several different goals, including promoting access and attainment of postsecondary education for qualified students; promoting student choice among public and private institutions; encouraging students to attend in-state rather than out-of-state institutions to reduce “brain drain;” and promoting and recognizing academic achievement in high school or college (Doyle, 2008; Heller, 2004). Need-based aid programs are used to promote access, equalization funds are used to promote choice, and merit-based aid programs are used to reward academic achievement (Doyle, 2008).

States vary in the foci and magnitude of their grant aid. Two states award 100% of their aid for undergraduates based on financial need only (Rhode Island and Wyoming) while one state (Louisiana) awards virtually all aid to undergraduates based only on merit (National Association of State Scholarships and Grant Aid Programs [NASSGAP], 2007). Some states (e.g., South Carolina and Georgia) offer over $1,000 in grant aid per full-time equivalent (FTE) undergraduate student, while others (e.g., Hawaii and Wyoming) offer less than $10 per FTE (College Board, 2008a). State grant expenditures represent at least one-fifth of higher education operating expenses in South Carolina (33%), Vermont (23%), Georgia (22%), West Virginia (21%), New York (21%), Pennsylvania (20%), and Indiana (20%), but less than 1% of higher education operating expenses in North Dakota, Alabama, Arizona, Alaska, Hawaii, and Wyoming (NASSGAP, 2007).

States also vary in the ways they disseminate information about financial aid (McDonough et al., 2007), as well as the extent to which they provide other forms of support for students’ college enrollment. Future research should use these variations to examine the ways that “college access marketing” campaigns influence college-related behaviors (McDonough et al., 2007), as well as how the effects of financial aid on college opportunity vary based on the availability of other resources that promote college enrollment. Perhaps the most comprehensive state-grant program, Indiana’s Twenty-first Century Scholarship program, includes early commitment of aid, activities to involve parents in college planning and preparation, and academic and social support services for students. Based on a review of data from a survey of students and state records, St. John et al. (2002) found that program participants were more likely to enroll in public 2-year, public 4-year, private, and out-of-state colleges and universities than non-participants and to remain continuously enrolled during the first year of college. Some evidence suggests that, compared to non-participants, program participants are more likely to persist from the first to second semester of their freshman year of college and to attain an associate’s degree (St. John et al., 2005, 2002; St. John, Musoba, Simmons & Chung, 2004), while other research suggests that scholars are less likely than non-recipients to attain a bachelor’s degree within 6 years (St. John et al., 2008). Students who take the Scholars Pledge are more likely than other students to graduate from high school with an Honors diploma (which requires completion of an advanced preparatory curriculum), suggesting that bachelor’s degree attainment rates for Scholars may improve over time (St. John et al., 2008, 2002; St. John, 2003; St. John, Musoba, Simmons & Chung, 2004) conclude that the promise of the availability of financial aid for college during the 8th grade encouraged students to engage in behaviors required to prepare for college. Schultz and Mueller (2006) conclude that the available research provides “promising evidence” of the program’s effectiveness for participating students, as the available research shows positive outcomes but uses a quasi-experimental rather than an experimental design.

Future research might also explore how the effects on students’ college-related behaviors vary based on other aspects of financial aid, including the providers of information about aid and the extent to which use of financial aid is framed as the “norm” or “default.” For example, Long (2009) speculates that students’ use and knowledge of financial aid would be substantially greater if students were automatically informed of their eligibility (using IRS data for example) than when students must proactively complete the Federal Application for Federal Student Aid (FAFSA).

Recognize the Contribution of Multiple Theoretical Approaches

Finally, future research should recognize the potential contribution of multiple theoretical approaches for understanding the ways financial aid influences students’ college-enrollment behaviors. As described in the conceptual framework section, no one theoretical perspective alone is sufficient. As a substantial body of research demonstrates, human capital models are useful for understanding that financial aid promotes college enrollment by increasing the resources that students and families have to pay college prices. Nonetheless, human capital models alone are insufficient for explaining inconsistencies in students’ decision-making (Long, 2009) or for understanding how the relationship between student aid and enrollment-behaviors varies based on other aspects of context (Manski, 2009).

Other perspectives, including those from behavioral economics, psychology, and sociology, may be especially useful for understanding how students acquire financial aid information and develop perceptions about financial aid, how students use this information and their perceptions to make college-related decisions, and how and why these relationships may vary across groups.

Conclusion

Clearly, financial aid is a critical policy-lever for increasing postsecondary access and choice, especially for Blacks, Hispanics, low-income students, and other groups underrepresented in higher education. This chapter argues research that more explicitly examines the ways financial aid is mediated by the context in which students are situated will generate important insights not only for improving public policy and practice but also for demonstrating accountability in financial aid expenditures. This chapter also argues that no single theoretical perspective or methodological approach alone is sufficient for completely understanding all that is still unknown about the effects of financial aid on college enrollment for different groups of students. Multiple theoretical frameworks, using multiple methods, addressing multiple questions, focusing on multiple populations, and recognizing multiple dimensions of context will help to address these knowledge gaps, an important step toward maximizing the effectiveness of student financial aid programs.