Fewer People Taking Out Student Loans For School? I Doubt It! (Update)

Sallie Mae's annual How America Pays For College results are in for 2008-09 and are causing quite a stir:

• New York Times:  Survey Finds Many That Many Families Don't Borrow for College
• NPR:  Fewer Students Borrow for College.  Those Who Do Spend More.
• Chronicle of Higher Education:  Fewer than Half of Families Paid for Undergraduate Education With Loans This Past Year, Survey Finds

Here is a summary of their results:

"The study found that among American families, 51 percent received grants and scholarships, 25 percent of students borrowed federal loans, 12 percent of students borrowed private education loans and 5 percent used credit cards to pay for college expenses. Fifty-eight percent of families paid for college last year without borrowing at all."

If we needed any further proof that little rigor or analysis goes into deciding how much debt a student decides to take on, the survey found that few had any idea of translating that debt into monthly payments:

"How America Pays for College found that 58 percent of families who borrowed did not take the student’s expected starting salary into consideration when deciding whether or how much to borrow. In addition, 23 percent of students could not answer when asked to estimate their future monthly student loan payment. The remaining 77 percent gave estimated future student loan payments that showed little correlation with the total amounts the students said they borrowed."

So, why do I doubt the results? 

It has to do with data that the Dept. of Education released earlier this year about federal loan borrowing which showed federal loan volume rose by 20.8% in 2008-09 largely driven by an increase in the number of loans, which rose by 23.6%:

Market Share Analysis

Program	2008-09	2007-08
DL	26.1%	20.5%
FFELP	73.9%	79.5%

Growth (2008-09 vs. 2007-08)

	DL	FFELP	OVERALL
Schools	51.1%	-3.5%	6.9%
Loans	55.2%	14.3%	23.6%
Loan $	54.2%	12.2%	20.8%

Data provided by NASFAA News: 

Direct Loan volume as of Feb. 25, 2009 compared to last year's volume

	2008-09	2007-08
Schools	1,620	1,072
number of loans	4.5 million	2.9 million
total loan dollars	$20.2 billion	$13.1 billion

Same for FFEL as of Feb. 26 2009

	2008-09	2007-08
Schools	4,424	4,583
number of loans	11.2 million	9.8 million
total loan dollars	$57.1 billion	$50.9 billion

----------------------------------------

How to account for the fact that 3.0 million more loans and over $13.0 billion more in federal loan volume was taken out in 2008-09 and yet somehow the percentage of families who borrowed for college in the Sallie Mae study declined from 47% in in 2007-08 to 42% in 2008-09. 

• For those who think the contraction of private loans account for the difference, it seems unlikely that this could account for this.  Based on SLA's estimate of a $6-$7 billion decline in private loan volume for 2008-09, overall borrowing including both federal and private loans would still be up over $6-$7 billion or roughly 7%-8%. 
  
• Or maybe it's a jump in enrollment accounting for the increase in borrowing?  Nope.    National Center for Education Statistics projected that enrollment in degree-granting institutions would remain flat in 2008 as compared to 2007 at 18.2 million students...
  
• Or maybe the average loan size increased while the number of borrowers decreased.  Drats,  that didn't happen either.  In fact, the average loan declined by 2.3% to $4,924 in 2008-09.  

So, how do I account for the difference:  One word:  METHODOLOGY.  Here is how they describe it in their report: 

"For this year’s study, Gallup surveyed 800 college-going students and 804 parents of such students in March and April 2009 on a series of questions that gauged Americans’ willingness and ability to pay for postsecondary education."

Given a choice between a telephone survey of 1,604 people or data on 15.7 million federal loans, I would have to consider the latter a more reliable gauge of what actually happened in 2008-09.  Perhaps, Sallie Mae relied too heavily on contacting their own private loan borrowers who took out loans in 2007-08, since many of them apparently didn't qualify in 2008-09 as SLM's private loan originations fell by 37.1% over that period.   

Rather than rely on my analysis, I would love to hear from the financial aid administrators, or better yet, this seems well suited to an SLA Flash Survey.  Anyone agree? 

Let's see how long it takes someone in MSM to question the survey results.  The clock is ticking...

------------------------
Update (August 25, 2009)

Mark Kantrowitz, publisher of FinAid.org, just commented on the survey on the NY Times blog, "The Choice.  He identified some differences in methodology between the Sallie Mae study and the NPSAS data which help explain the differences in results:

"The Sallie Mae survey focused on the percentage of families borrowing (or not borrowing) for college in the current academic year. According to the 2007-08 National Postsecondary Student Aid Study (NPSAS), 53% of students at 4-year public colleges and 41% of students at 4-year private nonprofit colleges did not borrow from federal or private education loans in 2007-08. However, more students borrow at some time during their undergraduate education than in any given year. Among students graduating with Bachelor’s degrees in 2007-08, 38.9% of students at 4-year public colleges and 29.4% of students at 4-year private nonprofit colleges graduated with no debt. Overall, only 13.7% of Bachelor’s degree recipients at 4-year colleges (and 34.6% of Associate’s degree recipients at 2-year colleges) graduated with no debt.

There’s also another inconsistency between the Sallie Mae survey data and the NPSAS data. 45.3% of NPSAS students had AGI less than $35,000, compared with only 16.9% of Sallie Mae survey respondents. The two were similar for families with income between $35,000 and $50,000, 11.6% and 12.1%, respectively. But 43.0% of NPSAS respondents had income over $50,000, compared with 71.0% of Sallie Mae respondents. Likewise, 15.3% of NPSAS respondents had income over $100,000 compared with 33.5% of Sallie Mae respondents. So it would appear that the Sallie Mae survey is skewed toward higher income families (who don’t need to borrow as much) as compared with the NPSAS. Considering that the NPSAS surveyed 140,000 undergraduate students and the Sallie Mae survey only 1,600 students and parents, sample size may also have had an impact on the results."

Comments

[Craigie]

National loan volumes have little to do with what is happening at the individual level. National Center for Education Statistics studies have been consistent -- about one third of students borrow during a given academic year, but, for those students who complete a bachelors degree, well over 60 percent have borrowed at some point during their undergraduate careers. The nonscientific Sallie Mae survey is high, not low.

National loan volumes are also skewed by graduate/professional school borrowing. While graduate/professionals are a tiny percentage of borrowers, they borrow much more per person. Grad PLUS is replacing the use of alternative loans, thus increasing loan counts. A typical graduate/professional student who borrows will take out three Federal loans rather than two: Grad PLUS, Stafford Sub and Stafford Unsub, rather than just Stafford Sub and Stafford Unsub as in the old days.

In addition, in an economic downturn, people tend to go back to school. Pell Grants are way up. High school grads who weren't going to go to college may decide to do so finally. College drop-outs may give it another try. College grads may feel the need for a masters degree.

All the indicators point to the postsecondary universe of attendees expanding rather than proportion of usual college-goers' borrowing increasing. The credit card "survey" that Nellie Mae (now Sallie Mae) publishes every year ismuch-hyped in the MSM but has little relevance. In any case, its main purpose -- to persuade policymakers to increase Federal education loan limits -- has been fulfilled. Let's now see whether the analysis from Nellie/Sallie is correct: that credit card usage by college students will now plummet. Doubt it. College students were never using credit card in place of student loans; it was as myth, check NCES . . . In addition, it takes a long time for FFELP numbers to work through the multi-layer reporting process. There is a long tail on the FFELP originations, 2008-09 of $75 billion FFELP and $20 billion DL.

[Tim Ranzetta]

Craigie,

Thanks for your comments about my post. I just wanted to make a few points to follow-up on your comments:
1. The main point of my post was not the absolute percentage of borrowers but rather the survey's claim that the percentage of borrowers had declined on a year over year basis. The percentage of borrowers is obviously impacted by the sample, with NCES being the most exhaustive sample, as it includes part-time students in it, which make up almost 40% of their population. It is presumed that these part-time students have less of a need to borrow, so that would explain why NCES figures are so much lower than other studies which tend to focus more on full-time students.
2. Regarding graduate/professional schools, you note that they make up a "tiny percentage of borrowers" (240,000 loans out of 13.95 million loans for 2007-08) so the fact that they are taking out 3 loans instead of two doesn't move the dial in terms of increasing the loan count. Also, the new GradPLUS program took effect in July of 2006 so the impact of this shift to GradPLUS would seem to be lessening over time. In fact, Grad PLUS volumes grew by about $700 million in 2007-08 or 27% according to the Dept. of Ed. (2008-09 values are not available), which would account for less than 10% of the $13 billion increase in federal loans for 2008-09.
3. I certainly agree with your point that "the postsecondary universe of attendees [is] expanding." I question the premise that this is the only factor driving the increase in student loan volumes. Analyzing NCES statistics since 1975 indicate the largest year-over-year increase in enrollment was 4.3% in 2002.
4. Finally, the issue you raise about FFELP numbers working through the multi-layer reporting process would seem to be a consistent issue on year over year basis so not sure why this would matter since I am measuring the year over year delta when presumably both years include this "multi-layer reporting process."

I stand by my analysis that a higher percentage of students and families borrowed to go to college in 2008-09. With stock markets cratering, home equity evaporating and job losses mounting while overall student loan volumes (federal and private) rose almost 10% (based on my estimates), it is hard for me to argue otherwise.

[Ms. C. Cryn Johannsen]

Tim, thanks for posting on the NPR story.

I've been following your blog for a while. I'm also blogging on the student lending crisis. I'm a promotional writer and marketer for Robert Applebaum's Forgive Student Loan Debt movement (http://www.forgivestudentloandebt.com/)

I've added you to my blog list.

I'm glad you responded to this highly problematic survey.

Yours,
Cryn

My blog: http://alleducationmatters.blogspot.com/

[Craigie]

The NCES DAS is publicly-available. Users can select full time/full year if only that grouping is desired. Some NCES pre-fab charts already provide this. The difference is not as much as anticipated. Even if it were, who is to say that almost all the increase in loans is not coming from that 40% of the population? After all, according to NASFAA, NAICU, etc., there is not much that full time students can do to increase their federal borrowing, hence their advocacy over many years for increasing federal loan limits as well as adding new loan programs/loan types.

While it might be presumed that part-time students have less of a need to borrow, that need to borrow might have increased much faster than that of the full-time students. And if the typical college student in 2009 is a 31 year old PT student, then why simply omit the whole group from the analysis.

The economy may impact private lending but has little impact on federal (FFEL and DL) lending. Federal lending is insulated by layers of consumer entitlements, not to mention the need to file the FAFSA (except for parent PLUS). While the economic crunch may be unusual and significant, it simply cannot be said that recessions, crashing stock markets, skyrocketing unemployment and home equity declines have never occurred since 1966. And people have been waiting almost since 1966 for a catastrophe that has never occurred. Loan metrics, including defaults, often react opposite to events in the general economy. While we may see something different for 2010 when looking back from 2015, there is nothing quantitative available to say that the economy has impacted the federal student loan situation in 2009, never mind 2008 or 2007.

The alternative explanation which is not being explored by the loan community is that FFEL margins were quite comfortable as of March 2008 and, if not for a triple crisis in the international credit markets unrelated to the FFEL program, then FFEL would still be riding high. In fact, your own analysis, if it is correct that volumes have truly increased apples-to-apples, year-over-year, would indicate that ECASLA seems to have completely offset any impacts of the international credit crisis on FFEL lending. CCRAA had little-to-no effect. In other words, the 2007 subsidy cuts (which only applied to new lending) were clearly not "extreme"; they appear irrelevant statistically. They were apparently "just right," in that, in combination with ECASLA, they have allowed loan volume to increase significantly while at the same time trimming away selective borrower benefits in favor of a strong national borrower benefits program -- as befits what is under the law a national social welfare program. The days of a lender offering different benefits packages to different schools (probably to less than 1% of Title IV institutions?) should not be the subject of nostalgia but rather viewed as a temporary moment in the inducements era when lenders and guaranty agencies applied a small portion of taxpayer subsidy funds in an effort to woo DL schools into the FFEL program. Although subsidy levels were much higher prior to DL than during, say 2006, you didn't see "borrower benefits" competition in 1993. In almost all cases, FFEL borrowers were charged the maximum fees and interest rates permitted by law and regulation prior to the competition posed by DL.

The FFEL program of 2008-09 is nothing like the FFEL program of 2006-07. Despite the rapid (and unregulated) implementation of ECASLA, the significant program changes occurring right around the time when lenders, guarantors and schools were making award year decisions could have resulted in significant shifting of loan issuance dates, for example, from 2007-08 dates to 2008-09 dates, thus accounting for some of the apparent increase. In addition, not sure where you are saying year over year delta, but the rather extreme difference in FFEL program operation (including the involvement of a whole new set of players -- custodians, feds, bypassing guarantor step, etc.) would make year over year comparison a major challenge.

Graduate/professional students can have a significant impact on national dollar volumes if not loan counts.

Between the 2003-04 and 2007-08 NPSAS studies a number of schools basically eliminated the requirement of borrowing. One of the consumer groups has a web page listing these schools and their individual "loan-free" income parameters. When viewed within a universe of thousands of Title IV institutions, this probably did not "move the dial" downward very much, if even noticeably. Nevertheless, when considering that, by many measures, borrowing was stagnant or declining from 1999-2000 to 2003-2004 and from 2003-2004 to 2007-2008, could a subsequent increase in borrowing partially consist of pent-up demand?

It is difficult to have it both ways, to explain away flat volumes as simply a function of frozen loan limits and then say that increasing volumes do not relate to the recent increases in loan limits. The economic factors have come and gone, but schools have said borrowing (and also their tuition levels) were kept in check by frozen loan limits.

If you meant to quote Kantrovitz as saying, "only 13.7% of Bachelor’s degree recipients at 4-year for-profit colleges graduated with no debt," then this could be the key to the discrepancy. If you assume that Sallie Mae's lending has gradually become more and more oriented towards the for-profit schools, then, no matter what happens in the economy, their borrowing can't increase very much as all because, according to NPSAS 2007-08, almost all of them were already borrowing, and, of those who were borrowing, they were already borrowing high dollar amounts. On the other hand, at 4yr public institutions a much lower percentage of students were borrowing and those who borrowed were borrowing far more modest amounts. 4yr private nonprofits were somewhere in the middle. In any case, the students at 4yr public and 4yr private nonprofit institutions had more room to increase borrowing levels than the students at 4yr for-profit institutions.