Journal of Social Research & Policy, Vol. 5, Issue 1, July 2014
For-Profit Higher Education and Financial Aid: Are Differences Driven by Schools or Students? LAURA DAWSON ULLRICH1 Department of Accounting, Finance and Economics Winthrop University
EMILY KATHERINE PRATT Center for Business and Economics Research University of Tennessee
Abstract Acquiring a college education has become increasingly important in maintaining a competitive edge in the job market over the past two decades; the number of colleges and universities in the United States has grown right along with higher education’s importance. While traditional four-year and community colleges are expanding and multiplying, so are proprietary, for-profit colleges and universities. It has been noted that the next “bubble” to burst could be in student debt, which is cause for concern for students, their families and taxpayers as well. At the moment, proprietary schools receive federal aid and loans through taxpayer supported programs and social institutions, as well as private lending institutions. In order to determine to nature of borrowing differences, we use a two-stage probit model. The National Center for Education Statistics Beginning Postsecondary Longitudinal Study restricted-use data set provided a sample of 7620 students followed between 2003 and 2006. Results show that, after controlling for demographic, social and academic characteristics in the first stage, for-profit attendance still significantly increases student debt burdens. Federal loan debt is increased by $945 and total debt by $1495 after holding constant those contributing factors from the first stage. $945 represents a 48.2 percent increase from a mean federal loan debt of $1961. Thus, analysis shows that something specific to for-profit institutions has led to increased student debt burdens. Keywords: Education; Higher Education; Public Finance; Social Policy; Government Expenditures.
Introduction The terms “for-profit” and “higher education” do not seem to be at absolute odds upon first glance. Many current college students likely believe that their institution turns a yearly profit; after all, paying for college tuition is increasingly a struggle and includes private and public loans, work-study and scholarship applications. In 2012 the average student at a four-year public institution paid 32 percent higher tuition just ten years ago (National Center for Education Statistics, 2012). One might think the logical outcome of such a situation would be an increase in competitive, lower-cost institutions like community colleges and vocational schools, or even more students choosing to forgo college entirely. Yet college degree conferment has increased alongside the aforementioned tuition increases and data show a 160 percent increase in the number of students attending oftentimes pricier for-profit institutions
1
Postal Address: Department of Accounting, Finance and Economics, Winthrop University, 324 Thurmond Building, Rock Hill, SC 29733, USA. E-mail Address:
[email protected]
54 | JSRP
Laura Dawson Ullrich, Emily Katherine Pratt
over the past decade (Miller, 2011). Since 1980, for-profit higher education institutions have grown at nearly nine times the rate of their not-for-profit counterparts (Wilson, 2010). Not long ago, for-profit education was mostly restricted to small certificate-granting institutions and family-owned cosmetology schools. Over the past ten years, however, for-profit institutions have made their presence known via flashy billboards, television commercials spots and glossy ads, much of which is paid for using money from students’ federal financial aid (Wilson, 2010). Forprofits have moved out of their small locations and into skyscrapers and even scenic (and formerly financially-strapped) college campuses, and have begun conferring degrees previously restricted to more typical college environments. In 1996, only 5 percent of undergraduate students and 14 percent of those seeking vocational education chose to enroll in for-profits, mostly for said vocational training (Apling, 1993). This number had increased to 11 percent of undergraduate students by the year 2010. Over forty percent of about three thousand for-profit institutions are publicly traded and business is booming; this begs the question—where is the money coming from, and who is paying for for-profits’ breakneck growth? The money trail at for-profits seems less complicated than with not-for profit higher education. This is because many for-profits use federal aid up to the legal cusp (Miller, 2011). As the number of students attending for-profits has increased, so has the pressure on those institutions, particularly the more expensive ones, to make them Title IV eligible. Title IV of the Higher Education Act of 1965 created the Direct Loan and Pell Grant programs. On average, for-profit students (and thus their schools as well) get more out of the Pell grant program, to the tune of two hundred dollars more per student. Pell recipients may be awarded up to $5,645 per academic year, and award size is proportional to cost of attendance. Thus, students attending higher-cost for-profits receive larger awards than those studying at relatively inexpensive community colleges (Cellini, 2010; Deming, Goldin & Katz, 2012). The “90/10 Rule” permits schools to receive up to ninety percent of revenue from Title IV funding. Many for-profit institutions have chosen to take full advantage of this rule, to the tune of $26 billion per year (Miller, 2011). According to a report issued by the Department of Education in 2012, 29 for-profit institutions actually failed the 90/10 test that year, receiving more than 90 percent of their revenue from federal sources.2 Table 1 presents the five largest recipients of federal student aid in the United States during the 2011-2012 school year. All five are for-profit institutions. The table includes the total amount received in federal financial aid as well as the percentage of their revenue this represents. The University of Phoenix, now the largest higher education institution in the United States, receives over $4 billion in federal financial aid each year, which represents greater than 86 percent of their revenue. It is also interesting to note that out of the 2,057 for-profit institutions included in the 2012 report only 214 received less than 50 percent of their revenue from federal sources. This percentage is much, much higher than would be seen at not-for-profit institutions, who generally rely on a combination of tuition, fees, state aid, federal aid and grant dollars to support its revenue requirements. Table 1: For-Profit use of Federal Financial Aid Federal Financial Aid Percent of Total Institution Revenue from Fiscal Year Institution Revenues 2011-2012 University of Phoenix $4,050,140,993 86.09% Ashford University $1,170,463,064 86.82% DeVry University $1,035,797,640 75.45% ITT Technical University $984,810,740 66.48% Walden University $785,534,370 77.87% Source: https://studentaid.ed.gov/about/data-center/school/proprietary (2011-2012 report) Pressure to provide these financial services to prospective students has led to speculation that some for-profit institutions have used dishonest (or even fraudulent) methods to increase enrollment and financial aid eligibility. In order to be eligible to disperse federal financial aid 2
https://studentaid.ed.gov/about/data-center/school/proprietary (2011-2012 report)
For-Profit Higher Education and Financial Aid
55 | JSRP
under Title IV, a school must be accredited by an appropriate accrediting agency, have legal permission to operate within a state and be approved by the Department of Education for participation in federal student aid (GAO, 2010). Of the aforementioned standards, accreditation benchmarks may be the most stringent; however, they can potentially be circumvented with proper aid from a consulting firm. A warning is issued by the accrediting agency before a school’s accreditation is revoked, giving it time to polish itself for the next round. More easily skirted are student requirements under Title IV. Once a school is qualified for Title IV dispersal, it must ensure its prospective and current students meet certain requirements to receive federal aid. It is in this area of regulation that the Government Accountability Office has found some of the most damning evidence against for-profit schools. The GAO conducted a small undercover operation involving fifteen different for-profit institutions, including both large, publicly-traded schools and the smaller certificate-issuing sort. The aim was to observe both marketing and financial aid practices under the same circumstances that normal citizens might encounter them in for-profits. To get a clear picture on financial aid practices, GAO visited each of the fifteen institutions with a “student” of high income and one with little to no income. The results were startling, yet potentially unsurprising given the large incentives involved. According to the GAO report, four of the fifteen colleges visited encouraged the undercover students to commit fraud, usually by having them add dependents or hide the fact that they had very substantial sums of money in savings, so that they could receive larger amounts of federal loans and even Pell Grants (GAO, 2009). In fact, wealthier students were told not to report a sum of $250,000 in savings in some cases. In one case a financial aid officer called and confirmed that she had made a change to the student’s FAFSA to eliminate the savings and qualify the student for aid. (GAO, 2009) Students who wish to receive federal financial aid must have a high school diploma, homeschool certificate, GED or, lacking those, must pass a test called the Ability to Benefit test (ATB). If this requirement is fulfilled, prospective students may complete the Free Application for Federal Student Aid. In some cases, for-profits helped GAO “students” acquire fake diplomas so that they could then be eligible for federal financial assistance. The Department of Education plays a small role in the testing process. It merely certifies test-publishing companies, which are then charged with overseeing the process in each school at which their ATB tests are administered (GAO, 2010) The GAO uncovered multiple instances of students being coached during the testing process as well as school officials changing test answers after students had finished. There were many red flags raised by the GAO study, and further investigation is surely ongoing regarding these issues. Knowledge concerning the payments for-profits receive is important. Perhaps just as interesting, and less commonly researched, is the type of person who chooses to attended forprofits. For-profits often market themselves as a second chance for the middle-aged businessman who never received his Bachelor’s or a flexible option for better future pay for homemakers and single mothers. In short, they generally try to reach an audience the typical not-for-profit 4-year college finds nearly unattainable: the working or non-traditional student with bills to pay, a family to feed and oftentimes very little money to spare for an education. This paper will investigate both the choice to attend a for-profit institution as well as the levels of financial aid that attending students use while in school. We attempt to describe empirically the elements which factor in to for-profit attendance as well as the amount of aid utilized. Forprofits have not yet taken majority hold of traditional two- and four-year degrees, and despite a few offering Master’s and Doctoral programs, traditional colleges still reign supreme in those areas. Public not-for-profits fare better in enrollment in all the aforementioned levels than do their private and for-profit counterparts. However, for-profits enroll almost seventy-five percent of non-degree-seeking students, and are gaining ground on not-for-profit institutions each year (Kinser, 2009). Whereas minority students make up a third of students in public and private institutions, they make up one half of for-profit attendees (Kinser, 2009; Deming, Goldin & Katz, 2012). At first, the statistics paint a picture of better access for otherwise disadvantaged students (be it due to socioeconomic or time constraint issues), but this access may not be provided based on
56 | JSRP
Laura Dawson Ullrich, Emily Katherine Pratt
altruism alone. Typically, lower-income groups bring in the Pell grants and easy Direct Loans that provide such a large percentage of for-profit education’s revenues. As a for-profit enterprise, they may have a more vested interest in the entrance of new students than the success or financial stability of its more vulnerable student populations.
Review of Literature The literature on this topic is somewhat sparse, as it has only recently become an important issue of interest in higher education. However, some significant work has been done on topics similar to those we investigate. Deming, Goldin & Katz (2012) examine the types of students enrolling at for-profit institutions and how for-profit attendance impacts post-graduation outcomes. They focus on unemployment and “idleness” rates, and find that for-profit students end up with higher unemployment and “idleness” rates six years after enrollment. They also conclude that students in for-profit institutions have higher student debt loads and higher loan default rates. They conclude that some for-profits are doing a good job of educating and graduating students, but others are simply “agile predators”. The earliest assessment of characteristics of both for-profit attendees and the schools themselves comes from Apling’s study (1993) using the National Center for Education Statistics Integrated Post-Secondary Education Data System from 1988 (IPEDS88) and National Post-Secondary Aid Survey. Apling’s look at for-profit’s composition reveals that much has changed and much has remained the same in the “business” of education. In 1993 it was estimated that one quarter of Pell grants and one third of Stafford Loans went to for-profit students (Apling, 1993). The same holds true today. The difference is that the sheer volume of all types of student debt has grown, making the direct loan and Pell programs more expensive to finance and administer. Fuller, Manski & Wise (1980) provide early work on college choice patterns and the role of federal financial aid, in particular the Pell Grant, on matriculation. An overall 21 percent increase in enrollment was caused by the introduction of the Pell Grant, as estimated, with two times that increase enrollment associated with students who came from low-income backgrounds. Higher income groups experienced a mere 3 percent increase in enrollment associated with the Pell Grant while middle-class enrollment increased by 12 percent; thus the Pell Grant encouraged college entry greatly in those groups which may have had trouble affording college previously. In order to inspect the role of financial aid in student sub-baccalaureate entry, Cellini (2010) took a close look at a form of grant aid offered to students in California. The Cal Grant Program offers three different types of grant aid to three different types of student: Cal A and B both require certain high school grade point averages and cover tuition and fees, and just fees respectively, while Cal C is much less particular. The Cal C program covers only educational programs lasting between four months and 2 years, and also provides additional grant money to students who do not choose to attend a state community college. Also examined is the Pell Grant’s role in for-profit prevalence by county. Increases in California Cal C and Pell Grant money were strongly positively correlated with the creation of new for-profit institutions within California.
Data and Analysis Our analysis expands on the Deming, Goldin & Katz (2012) paper which finds that students have overall higher levels of student loan debt and higher default rates on the loans they receive. We expand on this analysis to investigate whether or not for-profit students are more likely to hold federal student loan debt than their not-for-profit counterparts, and whether for-profit attendance significantly impacts the amount of federal loan debt a student has relative to their total debt held. For our analysis we utilized the National Center for Education Statistics’ Beginning Post-Secondary 03/06 and 04/09 Longitudinal Study dataset (BPS). The 04/09 wave of this survey was also used by Deming, Goldin & Katz (2012) and contains a plethora of information on a sample of college
For-Profit Higher Education and Financial Aid
57 | JSRP
enrollees including high school achievement and employment. The BPS data is collected in regular intervals and we utilize both the 03/06 and 04/09 waves in order to compare time periods and see if there is a significant difference in results over time. As a first step, we utilized a probit model in order to investigate the factors that influence the probability that a student will attend a for profit institution. 3 Our hypothesis was that demographic factors (such as race, gender, and income) played a significant role in the choice to attend a for-profit school rather than a not-for-profit institution. This first model is a simple probit model with for-profit college attendance used as the dependent variable. The probit model is defined as follows: Pr( Y=1 ∣ X1, X2,…X11)= Φ(β0+ β1X+…+ βnX) where Φ is the cumulative standard normal distribution function. Marginal effects allow the regression coefficients to take derivative form, meaning that a percent change in one of the coefficients leads to an associated change in probability of the binary dependent variable becoming one, or affirmative for choosing to attend a for-profit institution in this case. Demographic and economic variables (race, gender, income, single parenthood, parents’ education level, number of hours worked, urban locale, distance from college attending) as well as academic variables (high school GPA) were included as independent variables which might influence the decision to attend a for-profit institution. This predicts the probability that a student will attend a for-profit school based on all of these factors. The second model follows Deming, Goldin & Katz (2012) and utilizes a basic OLS regression that analyzes the effects of for-profit attendance along with many other control variables (including the factors included in the probit model) on federal financial aid utilized. As mentioned in the introduction, the use of such aid, and the low payback rates, are currently a major source of contention between for-profit institutions and the federal government. The OLS model is as follows: Yi= β0 +β1Xi1+….+βnXin+ εi The data contained in the BPS survey includes high school achievement level, demographic information, even volunteer activities in which students participated and employment type. A large portion of BPS draws from the National Postsecondary Student Aid Survey which is collected by the NCES as well. We utilize two different surveys, 2003/06 and 2004/09. For the 2003/06 BPS survey, students surveyed began the college entrance process in 2003 and were followed up three years later in 2006.4 For the 2004/09 survey, participants entered college in 2004 and were followed up with in 2006 and 2009. Because the 2004/09 survey covers a longer period of time it is certainly a richer dataset, but we include data/results for both sets of data as 5 a robustness check. It is important to note that BPS does not distinguish between small forprofit institutions, like those that provide cosmetology certificates or mechanic training, and the for-profits which have grown rampantly over the past ten to fifteen years— those that offer associate degrees all the way up to doctoral degrees in some cases, and oftentimes bear the same accreditation as do large public universities. The summary statistics along with variable descriptions are presented for the 2003/06 and 2004/09 datasets in Table 2.
3
We also tested a logit model and the results were entirely consistent with the probit results. In the end we chose to present the probit results in the paper, as we predict the underlying probability of for-profit attendance to have a relatively normal distribution rather than a logistic distribution. In this case, we believe it to be more appropriate to use the probit model. 4 Obviously most of the students will not have graduated by the last survey date of the analysis. Therefore our financial aid data from the 03/06 survey is simply an analysis of what they have borrowed thus far, not what they will borrow in sum. 5 When we first began the inquiry we only had 2003/06 data. The 2004/09 data was recently released, and we were granted a restricted use license to utilize the new data as well.
58 | JSRP
Laura Dawson Ullrich, Emily Katherine Pratt
Table 2: Summary Statistics Variable Total Federal Loans (in thousands) - total amount of Stafford, Perkins and PLUS loans Any Federal Loans - binary variable indicating any usage of federal loans Federal Ratio - ratio of federal aid to total aid AttendedFP - binary variable indicating for-profit (or proprietary) school attendance Locale - urbanization level (decreasing) of a student's locale (range 1 to 7) Distance - Distance from most recent institution 2006 (hundreds of miles) Stopouts - number of spells between enrollment Family Contribution - expected family contribution per FAFSA (in thousands) Age Gender HSGPA - high school grade point average College GPA - GPA as of last survey (in hudreth points) Transfers - number of transfers through the end of the survey Tuition Minus Grants - cost of tuition less federal grants (in thousands) Hours Worked - number of hours worked per week while enrolled AGI - household adjusted gross income (ten-thousands) Parent's Education - student's parent/guardian education level Black - binary variable indicating Black race Hispanic - binary variable indicating student is Hispanic Other Race - binary variable indicating Asian, Pacific Islander or other minority race Single Parent - binary variable for single parenthood
2003 - 2006
2004 - 2009
$9.93
$6.73
0.719 45.579
0.486 50.970
0.049
0.084
2.977
2.933
1.750 0.140
1.500 0.492
$11.48 18.490 0.583 296.123 312.638 0.228
$10.92 20.305 0.586 295.997 314.078 0.522
4.855
4.412
12.998 6.244 5.554 0.120 0.089
24.608 5.847 5.165 0.137 0.014
0.014 0.018
0.105 0.066
Out of the thousands of variables provided by BPS, demographic as well as achievement-related variables were chosen to determine their individual and relative impacts upon student choice between for-profit or proprietary schools and not-for-profits. While the dataset is vast, and contains a large number of variables, there is missing data in many cases, and the choice of variables used was partially determined by the availability of reliable, full data. If data were missing for a student, that student was eliminated from the database. This process left us with a large sample of 7,620 students for the 2004/09 wave and 3,460 for the 2003/06 wave. Students’ race, adjusted gross income, gender, and single-parenthood status were thought to contribute to college selection and were chosen from BPS for that reason. Additionally, differences in parent education level, hours worked per week, level of urbanization of students’ locality, and distance from nearest higher education institution were thought to contribute to the likelihood a student might attend a for-profit.6 Prior research has confirmed these variables as important predictors as well (Deming, Goldin & Katz, 2012). For-profits are found more densely in urban areas and have generally lower admissions standards than their not-for-profit counterparts. Though community colleges and some four year universities struggle to provide more flexible scheduling, funding and staffing issues prevent them from providing class times that for-profits do, whether in the classroom or online. The second part of the model utilizes achievement variables gathered during students’ postsecondary attendance period, as well as the variables contained in the probit analysis, in order to The variable LOCALE indicates the degree of urbanization in the student’s location. This variable ranges from 1 (large city) to 7 (rural). 6
For-Profit Higher Education and Financial Aid
59 | JSRP
analyze the amounts of federal loans utilized during the time frame analyzed. This is important as many of the same variables that determine college choice may also determine the level of financial aid needed, for example parents’ education level or single parent status. However, the amount of financial aid needed is also likely determined by variables that are college specific, such as college GPA and the number of times a student has transferred from one institution to another. For the 2003/06 dataset, this is only three years into their college journey, so many of the respondents had not yet graduated. For the 2004/09 dataset it is five years since college admittance, so many more of the students will have completed their college degree (or have quit college all together). Therefore, we naturally expect federal financial aid use to be much higher in the 2004/09 data. The number of college transfers, the number of stop outs which occurred during the time period, college gradepoint average and tuition costs were analyzed for their contribution to total amount of federal debt owed at the end of the survey periods. For-profit attendance is our key independent variable as we wish to analyze what impact attending a for-profit college has on federal financial aid after holding academic, demographic and financial variables constant. We utilized two different dependent variables for the OLS analysis: total federal loans utilized and the ratio of federal financial aid to total aid. We hypothesize that both of these will be significantly and positively correlated with forprofit attendance even after holding other factors constant.7
Results The first analysis regarding the probability of for-profit attendance is shown in Table 3. The results presented are the marginal effects associated with the probit model. This analysis sheds light on some of the factors which contribute to for-profit attendance and shows that the indicators of for-profit attendance are highly consistent across datasets. The results conform to our expectations in general, with the exception of Gender, which is insignificant. Prior to running the model, we believed that women would be more likely to attend for-profit schools, all else constant, because much of the advertising done by the schools appears anecdotally to be targeted towards women. Two-thirds of forprofit institution students are female, which is sixteen percent above the average for all types of higher education for 2003 (Francis, 2007). What then has such a significant impact on female attendance of for-profits if empirically gender does not? The answer may displayed prominently in some advertisements, which portray mothers somehow easily managing young children and working toward a Master’s degree simultaneously at home. In fact, our results show that being a single parent increases the probability of taking the for-profit route by between 4 and 6 percent, perhaps because they provide more online courses and more non-traditional time slots than their not-for-profit counterparts do. Table 3: Probit Model Results attendedFP 2003-2006 2004-2009 Parent's Education -0.003± -0.006± (0.000) (0.001) Gender 0.003 0.002 (0.003) (0.003) Single Parent 0.044‡ 0.062± (0.018) (0.009) HSGPA 0.012± -0.003± (0.002) (0.000) Hours Worked 0.001± 0.000± (0.000) (0.000) AGI -0.001* -0.004± (0.000) (0.000) Family Contribution -0.000 -0.005† (0.000) (0.000) 7
In the future we plan to further analyze the impact on TOTAL debt (federal + private + other) in order to examine the impact on the person attending the school as well as the government.
Laura Dawson Ullrich, Emily Katherine Pratt
60 | JSRP
Age
0.008± (0.002) Distance 0.003± (0.001) Locale -0.010± (0.001) Black -0.004 (0.004) Hispanic 0.023± (0.007) Other Race 0.001 (0.006) ± Indicates significance at the 1% level ‡ Indicates significance at the 5% level † Indicates significance at the 10% level
0.001± (0.000) -0.001‡ (0.000) -0.017± (0.001) 0.009‡ (0.004) 0.025± (0.006) -0.010‡ (0.004)
In general, the results indicate that older, working, minority, urban students are more likely to attend for-profit institutions than their counterparts. In addition, students whose parents have higher incomes and education levels are significantly less likely to attend for-profit schools than those with lower incomes. Interestingly, race appears to be a more significant determinant of for-profit attendance in the 2004/09 dataset than it was in the 2003/06 dataset. According to the 2004/09 data, Black and Hispanic students are statistically more likely to attend for-profit schools than White students (all else equal), while students of other races are less likely to attend than whites. These results conform both to our expectations as well as the information provided by for-profit institutions, many of which state their main objective as educating non-traditional students in a more flexible environment. The second set of models examines the level of federal loans borrowed during the survey period. These results are presented in Table 4. Two dependent variables were utilized: total federal loan balance and a binary variable indicating whether or not federal loans were used by the student.8 The results indicate in general that, all else equal, for profit students are more likely to utilize federal loans, 26.3 percent more likely during 2003/06 and 13.4 percent more likely in 2004/09. In addition, during the 2003-2006 survey, for-profit attendance had a positive, and significant, impact on the level of federal borrowing, with for-profit students borrowing over $1,400 more than not-for-profit students, all else equal.
AttendedFP Locale Distance Stopouts Family Contribution Age
8
Table 4: Baseline Models 2003-2006 2004-2009 Any Any Total Federal Federal Total Federal Federal Loans Loans Loans Loans 1470.323‡ 0.263± 354.787 0.134‡ (655.891) (0.015) (353.844) (0.019) -105.040 0.011± -50.083 -0.003 (80.617) (0.003) (56.853) (0.003) 55.705 0.001 -2.540 0.000 (35.971) (0.001) (14.500) (0.001) -2740.419± -0.008± -1174.016± -0.019‡ (312.151) (0.016) (129.028) (0.007) -194.422± -0.078± -97.742± -0.007‡ (25.281) (0.000) (9.394) (0.000) -628.443± -0.026± -55.064± 0.008‡ (139.526) (0.007) (17.358) (0.000)
In other words, this variable has a value of 1 if federal loan balances > $0, 0 otherwise.
For-Profit Higher Education and Financial Aid
Gender
-140.174 (307.580) HS GPA 4.379± (1.322) College GPA -8.359± (2.394) Transfers -730.181± (250.618) Tuition Minus 596.908± Grants (430.607) Hours Worked -12.943 (10.452) AGI 419.027± (90.793) Parent's Education -193.564± (60.726) Black 1265.472± (417.410) Hispanic -1014.317 (476.848) Other Race 861.391‡ (1472.613) Single Parent -2139.943± (2821.221) ± Indicates significance at the 1% level ‡ Indicates significance at the 5% level † Indicates significance at the 10% level
-0.009 (0.011) 0.013‡ (0.006) -0.001± (0.000) 0.099± (0.013) 0.012± (0.000) 0.001† (0.000) 0.008± (0.000) -0.015± (0.002) 0.056± (0.017) -0.056± (0.021) 0.031 (0.045) 0.090± (0.048)
-501.015‡ (205.282) 3.960± (1.266) -9.870± (1.651) 2064.943± (162.605) 61.747± (18.805) -41.715± (8.291) 74.432‡ (34.174) -49.721 (43.604) 3183.504± (358.726) -530.119† (319.718) -1117.834± (431.949) -433.588 (431.949)
61 | JSRP
0.009 (0.010) -0.000 (0.000) -0.001‡ (0.000) 0.121‡ (0.007) 0.001 (0.000) -0.002‡ (0.004) 0.005‡ (0.000) -0.005‡ (0.002) 0.114‡ (0.015) -0.056‡ (0.016) -0.064‡ (0.018) -0.022‡ (0.021)
Results also indicate that students with lower family contributions utilized federal loans more frequently, and at higher rates. During the 2004/09 period, an additional $1,000 in family contributions decreased federal loan totals by $97 over what would have been expected otherwise. Race also appears to play a significant factor in federal loan utilization and balances. Black students are more likely than White students to borrow via federal loans, all else equal, and have significant higher balances. Hispanic students are less likely to borrow and have lower balances than White students, all else equal. One result that was consistent in the models presented in Table 4 was very surprising to us. The models indicate that students with higher family adjusted gross incomes are more likely to borrow from the federal government and have higher balances, all else equal. We believe that this may be due to the PLUS loans, which are issued to parents rather than students, since parents must have favorable credit scores to obtain such loans. We expect each of these demographic and economic variables to play a role in the level of debt a student must bear to obtain a college education. However, potentially most disquieting is the effect that simply attending a for-profit alone has on student debt burdens. In theory, the for-profit status of a school should have no impact on the level of debt a student will incur, holding all else constant. However, based on the 2003/06 dataset, all income- and background-based risk factors aside, attending a for-profit has the effect of increasing federal student loan burdens by $1,470. This represents a 14.8 percent increase when measuring from mean federal debt levels in the 2003/06 survey. One final set of models was examined using another dependent variable, the ratio of federal loans to total aid multiplied by 100. This variable indicates the student’s reliance on federal loans as compared to reliance on private loans, public and private grants and merit aid. The variable basically represents what percentage of their total aid comes from the federal government. The mean level of the ratio for the 2003/06 dataset was 45.58 and rose to 50.97 in 2004/09. The increase in federal reliance in the 2004/09 dataset may be an indicator of the tightening of private credit markets in the later part of the
62 | JSRP
Laura Dawson Ullrich, Emily Katherine Pratt
time period analyzed. A basic OLS model was estimated in order to test the determinants of federal aid reliance. The results of this model for both datasets are included in Table 5. Table 5: Federal Ratio Models Dependent Variable: Federal Ratio 2003-2006 AttendedFP 20.189± (1.622) Locale 0.881± (0.222) Distance -0.227± (0.074) Stopouts -0.475 (1.180) Family Contribution -0.388± (0.047) Age 0.499 (0.516) Gender 1.100 (0.798) HS GPA -4.463± (0.416) College GPA -0.060± (0.007) Transfers 0.721 (0.864) Tuition Minus Grants 0.888± (0.075) Hours Worked 0.125± (0.030) AGI -0.578± (0.154) Parent's Education -1.392± (1.690) Black 10.693± (1.229) Hispanic 3.043‡ (1.368) Other Race 7.185‡ (3.185) Single Parent 16.726± (2.654) ± Indicates significance at the 1% level ‡ Indicates significance at the 5% level † Indicates significance at the 10% level
2004-2009 21.095± (1.169) 0.219 (0.212) -0.320± (0.072) -0.238 (0.580) -0.398± (0.054) 0.295± (0.085) 1.872 (0.766) -3.033± (0.005) -0.050± (0.007) 1.082‡ (0.544) 0.829± (0.000) 0.148± (0.032) -0.999± (0.167) -1.087± (0.155) 6.863± (1.034) 3.600± (1.191) 0.394 (1.420) 12.013± (1.451)
The results indicate that for-profit school attendance is a positive and highly significant determinant of federal aid reliance in both datasets analyzed. All else equal, a student attending a for-profit institution relies on the federal government for 21 percentage points more aid than they would if they attended a not-for-profit school. This result may give credence to allegations that for-profit institutions encourage students to maximize their federal aid beyond what would be done at not-for-profit schools. As the GAO report showed, this may be done via misleading or even fraudulent practices.
For-Profit Higher Education and Financial Aid
63 | JSRP
Not surprisingly, students with higher high school GPAs are less reliant on federal aid, as they are likely eligible for larger amounts of merit aid as well as grants from state lotteries and other programs. Also, family contribution is negatively correlated with federal aid reliance. This may be due to the fact that families that are able to contribute higher amounts are generally ineligible for most federal aid programs, and may have greater access to private lending opportunities.
Conclusion All of our baseline results indicate that attending for-profit schools leads to higher reliance on federal financial aid than would be expected otherwise and may also lead to higher federal debt balances. The consistency of the results presented indicate that indeed something may be ‘different’ at for-profit schools, and the information gathered by the GAO, coupled with results such as this, may lead to additional questions for for-profit operators. The impact the “90/10” rule has on the world of financial aid has not gone unnoticed by the current presidential administration, and in turn, the stock market. Upon the release of GAO investigation details and promises from the Department of Education that a thorough study of for-profit higher education would follow, stock prices for the major publicly traded for-profits fell by as much as 6 percent (Laurerman & Hechinger, 2010). It is of little wonder, statistically, why the government and public are concerned. Between 2000 and 2010, the number of students enrolled at for-profits grew 160 percent (Miller, 2011). In 2010, the Department of Education estimated that 26 percent of all federal student loans and 43 percent of all loans in default are attributable to for-profit education (Department of Education, 2011). The primary incentive to work toward a higher-level degree is generally the increased probability that one will earn larger wages than with a high school diploma alone. While a year or so of schooling past high school may help, generally the degree itself is important for securing higher wages. The problem amongst many in the US who begin the college journey is that they often do not finish with that all-important document in hand – the diploma. They do, however, receive many other less auspicious documents from college — loan statements. What then is the fate of the student (in reality a GAO employee), who wanted to pursue a Bachelor’s in Business Administration at an expensive, publicly-traded for-profit with a staggeringly low 9 percent graduation rate (GAO, 2010)? While for-profits might blame the population of students that choose to attend their schools, the above analysis suggests that there may be other factors, for-profit specific factors, that are leading to reduced outcomes and higher debt balances. The Department of Education has proposed and implemented different reforms to ameliorate the three main problems associated with for-profits: asymmetric information, drain on federal financial aid and fraud. Students are inherently at an informational disadvantage when it comes to navigating financial aid, leaving them vulnerable to predatory marketing, questionable financial aid advice and, as the GAO’s investigation very clearly indicated, outright fraud. In turn, for-profits are able to use the student as an intermediary to quick and easy profit via the federal financial aid system; aside from a diploma, GED or the Ability to Benefit test (around which schools have been known to maneuver), few barriers exist. Regulation reform was introduced by the Department of Education under Arne Duncan and enforcement began July 1, 2011. Most important in increasing information and allowing potential students to make sound financial decisions is the requirement that schools publish graduation and job placement rates for each program they offer ( Department of Education, 2011). In addition, schools with low federal debt repayment rates now have to issue warnings of such on their websites and in all admission materials. The question is, will students take note and substitute to not-for-profit institutions or will for-profits get more creative with their admission techniques? Enrollment fell by 3% at for-profit institutions in 2012, but until full data is available, it is not possible to know whether or not the new regulations drove the declining enrollment.9 Future waves of BPS data will help to answer these, and other, questions surrounding for-profit education in the United States. 9
http://online.wsj.com/news/articles/SB10001424052970203937004578076942611172654
64 | JSRP
Laura Dawson Ullrich, Emily Katherine Pratt
References 1.
Apling, R.N. (1993). Proprietary Schools and Their Students. Journal of Higher Education, 64(4), pp. 379-414. http://dx.doi.org/10.2307/2960049
2.
Cellini, S.R. (2010). Financial Aid and For-Profit Colleges: Does Aid Encourage Entry?. Journal of Policy Analysis and Management, 29(3), pp. 526-552. http://dx.doi.org/10.1002/pam.20508
3.
Deming, D., Goldin, C., & Katz, L. (2012). The For-Profit Postsecondary School Sector: Nimble Critters or Agile Predators?. Journal of Economic Perspectives, 26(1), pp. 139 – 164. http://dx.doi.org/10.1257/jep.26.1.139
4.
Department of Education. Press. (2010). Department of Education Establishes New Student Aid Rules to Protect Borrowers and Taxpayers. Retrieved March 11, 2012, from http://www.ed.gov/news/press-releases/department-education-establishes-newstudent-aid-rules-protect-borrowers-and-tax.
5.
Francis, D.R. (2007). Why Do Women Outnumber Men in College? The National Bureau of Economic Research. Retrieved February 3, 2013, from http://www.nber.org/digest/jan07/w12139.html.
6.
Fuller, W.C., Manski, C.F., & D.A. Wise. (1980). The Impact of the Basic Educational Opportunity Grant Program on College Enrollments. Cambridge, MA; John Fitzgerald Kennedy School of Government, Harvard University.
7.
Kinser, K. (2009). How the For-Profit Sector Contributes to Access in Higher Education. Enrollment Management Journal, 3(4), pp 23-44.
8.
Lauerman, L., & Hechinger, J. (2010) Education's Duncan Cracks Down on For-Profit Colleges. Bloomberg - Business & Financial News, Breaking News Headlines. Retrieved March 10, 2013 from http://www.bloomberg.com/news/2010-0813/duncan-says-education-department-to-step-up-for-profit-college-enforcement.html
9.
Miller, B. (2011). Are You Gainfully Employed? Setting Standards for For-Profit Degrees. Education Sector Reports. Washington, D.C. Retrieved February 1, 2013 from http://www.educationsector.org/sites/default/files/publications/Gainful-Report_RELEASE.pdf.
10. National Center for Education Statistics (NCES). Fast Facts. Retrieved February 1, 2013 from http://nces.ed.gov/fastfacts/index.asp?faq=FFOption6#faqFFOption6.
(2012).
11. United States. Department of Education. (2009). Program Integrity: Gainful Employment Notice of Proposed Rulemaking. Retrieved February 12, 2013 from http://www2.ed.gov/policy/highered/reg/hearulemaking/2009/ge-faq.pdf . 12. United States. Government Accountability Office (GAO). (2010) For-profit Colleges: Undercover Testing Finds Colleges Encouraged Fraud and Engaged in Deceptive and Questionable Marketing Practices. Testimony before the Committee on Health, Education, Labor, and Pensions, U.S. Senate. U.S. Government Accountability Office. Retrieved February 1, 2013, from http://www.gao.gov/new.items/d10948t.pdf.
For-Profit Higher Education and Financial Aid
65 | JSRP
13. United States. Government Accountability Office (GAO). (2009). Stronger Department of Education Oversight Needed to Help Ensure Only Eligible Students Receive Federal Financial Aid. (testimony of George A. Scott, Director Education, Workforce and Income Security). Retrieved March 1, 2013, from http://www.gao.gov/products/GAO-09-600. 14. Wilson, R. (2010). For-Profit Colleges Change Higher Education's Landscape. The Chronicle of Higher Education. Retrieved February 1, 2013 from http://chronicle.com/article/For-Profit-Colleges-Change/64012/ .
